Cohen et al_CLIA CRC_Clinical analysis 11022024
library(swimplot) library(pheatmap) library(reshape2) library(coxphf) library(grid) library(gtable) library(readr) library(mosaic) library(dplyr) library(survival) library(broom) library(survminer) library(ggplot2) library(scales) library(coxphf) library(ggthemes) library(tidyverse) library(gtsummary) library(flextable) library(parameters) library(car) library(ComplexHeatmap) library(tidyverse) library(readxl) library(survival) library(janitor) library(openxlsx) library(writexl) library(rms) library(DT)
#ctDNA Detection rate by Stage and Window
#MRD Window
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"))
circ_data$Stage <- factor(circ_data$Stage, levels=c("I","II","III"))
circ_data <- subset(circ_data, ctDNA.MRD %in% c("NEGATIVE", "POSITIVE"))
positive_counts_by_stage <- aggregate(circ_data$ctDNA.MRD == "POSITIVE", by=list(circ_data$Stage), FUN=sum)
total_counts_by_stage <- aggregate(circ_data$ctDNA.MRD, by=list(circ_data$Stage), FUN=length)
combined_data <- data.frame(
Stage = total_counts_by_stage$Group.1,
Total_Count = total_counts_by_stage$x,
Positive_Count = positive_counts_by_stage$x,
Rate = (positive_counts_by_stage$x / total_counts_by_stage$x) * 100 # Convert to percentage
)
combined_data$Rate <- sprintf("%.2f%%", combined_data$Rate)
overall_total_count <- nrow(circ_data)
overall_positive_count <- nrow(circ_data[circ_data$ctDNA.MRD == "POSITIVE",])
overall_positivity_rate <- (overall_positive_count / overall_total_count) * 100 # Convert to percentage
overall_row <- data.frame(
Stage = "Overall",
Total_Count = overall_total_count,
Positive_Count = overall_positive_count,
Rate = sprintf("%.2f%%", overall_positivity_rate)
)
combined_data <- rbind(combined_data, overall_row)
print(combined_data)
#Surveillance Window
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data$ctDNA.Surveillance <- factor(circ_data$ctDNA.Surveillance, levels=c("NEGATIVE","POSITIVE"))
circ_data$Stage <- factor(circ_data$Stage, levels=c("I","II","III"))
circ_data <- subset(circ_data, ctDNA.Surveillance %in% c("NEGATIVE", "POSITIVE"))
positive_counts_by_stage <- aggregate(circ_data$ctDNA.Surveillance == "POSITIVE", by=list(circ_data$Stage), FUN=sum)
total_counts_by_stage <- aggregate(circ_data$ctDNA.Surveillance, by=list(circ_data$Stage), FUN=length)
combined_data <- data.frame(
Stage = total_counts_by_stage$Group.1,
Total_Count = total_counts_by_stage$x,
Positive_Count = positive_counts_by_stage$x,
Rate = (positive_counts_by_stage$x / total_counts_by_stage$x) * 100 # Convert to percentage
)
combined_data$Rate <- sprintf("%.2f%%", combined_data$Rate)
overall_total_count <- nrow(circ_data)
overall_positive_count <- nrow(circ_data[circ_data$ctDNA.Surveillance == "POSITIVE",])
overall_positivity_rate <- (overall_positive_count / overall_total_count) * 100 # Convert to percentage
overall_row <- data.frame(
Stage = "Overall",
Total_Count = overall_total_count,
Positive_Count = overall_positive_count,
Rate = sprintf("%.2f%%", overall_positivity_rate)
)
combined_data <- rbind(combined_data, overall_row)
print(combined_data)
#Anytime post-surgery
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data$ctDNA.anytime <- factor(circ_data$ctDNA.anytime, levels=c("NEGATIVE","POSITIVE"))
circ_data$Stage <- factor(circ_data$Stage, levels=c("I","II","III"))
circ_data <- subset(circ_data, ctDNA.anytime %in% c("NEGATIVE", "POSITIVE"))
positive_counts_by_stage <- aggregate(circ_data$ctDNA.anytime == "POSITIVE", by=list(circ_data$Stage), FUN=sum)
total_counts_by_stage <- aggregate(circ_data$ctDNA.anytime, by=list(circ_data$Stage), FUN=length)
combined_data <- data.frame(
Stage = total_counts_by_stage$Group.1,
Total_Count = total_counts_by_stage$x,
Positive_Count = positive_counts_by_stage$x,
Rate = (positive_counts_by_stage$x / total_counts_by_stage$x) * 100 # Convert to percentage
)
combined_data$Rate <- sprintf("%.2f%%", combined_data$Rate)
overall_total_count <- nrow(circ_data)
overall_positive_count <- nrow(circ_data[circ_data$ctDNA.anytime == "POSITIVE",])
overall_positivity_rate <- (overall_positive_count / overall_total_count) * 100 # Convert to percentage
overall_row <- data.frame(
Stage = "Overall",
Total_Count = overall_total_count,
Positive_Count = overall_positive_count,
Rate = sprintf("%.2f%%", overall_positivity_rate)
)
combined_data <- rbind(combined_data, overall_row)
print(combined_data)#ctDNA MRD Detection rate Stage I/II vs III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels = c("NEGATIVE", "POSITIVE"))
circ_data$Stage_Grouped <- factor(ifelse(circ_data$Stage %in% c("I", "II"), "I/II", "III"))
contingency_table <- table(circ_data$Stage_Grouped, circ_data$ctDNA.MRD)
chi_square_test <- chisq.test(contingency_table)
print(contingency_table)
NEGATIVE POSITIVE
I/II 162 15
III 172 56print(chi_square_test)
Pearson's Chi-squared test with Yates' continuity correction
data: contingency_table
X-squared = 16.741, df = 1, p-value = 4.285e-05#Demographics Table
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data_subset <- circ_data %>%
select(
Age,
Gender,
PrimSite,
pT,
pN,
Stage,
Grade,
NAC,
ACT,
MSI,
BRAF.V600E,
RAS,
DFS.Event,
OS.months) %>%
mutate(
Age = as.numeric(Age),
Gender = factor(Gender, levels = c("Male", "Female")),
PrimSite = factor(PrimSite, levels = c("Right-sided colon", "Left-sided colon")),
pT = factor(pT, levels = c("T0","T1-T2", "T3-T4")),
pN = factor(pN, levels = c("N0", "N1-N2")),
Stage = factor(Stage, levels = c("I","II", "III")),
Grade = factor(Grade, levels = c("G1", "G2", "G3","GX")),
NAC = factor(NAC, levels = c("TRUE", "FALSE"), labels = c("Neoadjuvant Chemotherapy", "Upfront Surgery")),
ACT = factor(ACT, levels = c("TRUE", "FALSE"), labels = c("Adjuvant Chemotherapy", "Observation")),
MSI = factor(MSI, levels = c("MSS", "MSI-High")),
BRAF.V600E = factor(BRAF.V600E, levels = c("WT", "MUT"), labels = c("BRAF WT", "BRAF V600E")),
RAS = factor(RAS, levels = c("WT", "MUT"), labels = c("RAS WT", "RAS Mut")),
DFS.Event = factor(DFS.Event, levels = c("TRUE", "FALSE"), labels = c("Recurrence", "No Recurrence")),
OS.months = as.numeric(OS.months))
table1 <- circ_data_subset %>%
tbl_summary(
statistic = list(
all_continuous() ~ "{median} ({min} - {max})",
all_categorical() ~ "{n} ({p}%)")) %>%
bold_labels()
table1| Characteristic | N = 7951 |
|---|---|
| Age | 61 (13 - 91) |
| Gender | |
| Male | 407 (51%) |
| Female | 388 (49%) |
| PrimSite | |
| Right-sided colon | 411 (52%) |
| Left-sided colon | 384 (48%) |
| pT | |
| T0 | 3 (0.4%) |
| T1-T2 | 133 (17%) |
| T3-T4 | 654 (83%) |
| Unknown | 5 |
| pN | |
| N0 | 310 (39%) |
| N1-N2 | 482 (61%) |
| Unknown | 3 |
| Stage | |
| I | 47 (5.9%) |
| II | 262 (33%) |
| III | 486 (61%) |
| Grade | |
| G1 | 84 (11%) |
| G2 | 559 (72%) |
| G3 | 127 (16%) |
| GX | 4 (0.5%) |
| Unknown | 21 |
| NAC | |
| Neoadjuvant Chemotherapy | 0 (0%) |
| Upfront Surgery | 795 (100%) |
| ACT | |
| Adjuvant Chemotherapy | 522 (66%) |
| Observation | 273 (34%) |
| MSI | |
| MSS | 664 (84%) |
| MSI-High | 131 (16%) |
| BRAF.V600E | |
| BRAF WT | 699 (88%) |
| BRAF V600E | 96 (12%) |
| RAS | |
| RAS WT | 459 (58%) |
| RAS Mut | 336 (42%) |
| DFS.Event | |
| Recurrence | 141 (18%) |
| No Recurrence | 654 (82%) |
| OS.months | 27 (0 - 103) |
| 1 Median (Range); n (%) | |
fit1 <- as_flex_table(
table1,
include = everything(),
return_calls = FALSE,
strip_md_bold = TRUE)Warning: The `strip_md_bold` argument of `as_flex_table()` is deprecated as of gtsummary 1.6.0.fit1Characteristic | N = 7951 |
|---|---|
Age | 61 (13 - 91) |
Gender | |
Male | 407 (51%) |
Female | 388 (49%) |
PrimSite | |
Right-sided colon | 411 (52%) |
Left-sided colon | 384 (48%) |
pT | |
T0 | 3 (0.4%) |
T1-T2 | 133 (17%) |
T3-T4 | 654 (83%) |
Unknown | 5 |
pN | |
N0 | 310 (39%) |
N1-N2 | 482 (61%) |
Unknown | 3 |
Stage | |
I | 47 (5.9%) |
II | 262 (33%) |
III | 486 (61%) |
Grade | |
G1 | 84 (11%) |
G2 | 559 (72%) |
G3 | 127 (16%) |
GX | 4 (0.5%) |
Unknown | 21 |
NAC | |
Neoadjuvant Chemotherapy | 0 (0%) |
Upfront Surgery | 795 (100%) |
ACT | |
Adjuvant Chemotherapy | 522 (66%) |
Observation | 273 (34%) |
MSI | |
MSS | 664 (84%) |
MSI-High | 131 (16%) |
BRAF.V600E | |
BRAF WT | 699 (88%) |
BRAF V600E | 96 (12%) |
RAS | |
RAS WT | 459 (58%) |
RAS Mut | 336 (42%) |
DFS.Event | |
Recurrence | 141 (18%) |
No Recurrence | 654 (82%) |
OS.months | 27 (0 - 103) |
1Median (Range); n (%) | |
save_as_docx(fit1, path= "~/Downloads/table1.docx")#Heatmap with Clinical & Genomics Factors
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data %>% arrange(Stage)
circ_datadf <- as.data.frame(circ_data)
ha <- HeatmapAnnotation(
Stage = circ_data$Stage,
Gender = circ_data$Gender,
PrimSite = circ_data$PrimSite,
pT = circ_data$pT,
pN = circ_data$pN,
Grade = circ_data$Grade,
ACT = circ_data$ACT,
MSI = circ_data$MSI,
BRAF.V600E = circ_data$BRAF.V600E,
RAS = circ_data$RAS,
ctDNA.MRD = circ_data$ctDNA.MRD,
ctDNA.Surveillance = circ_data$ctDNA.Surveillance,
ctDNA.anytime = circ_data$ctDNA.anytime,
DFS.Event = circ_data$DFS.Event,
col = list(Stage = c("I" = "seagreen1", "II" = "orange", "III" = "purple"),
Gender = c("Female" = "goldenrod" , "Male" = "blue4"),
PrimSite = c("Right-sided colon" = "brown", "Left-sided colon" ="darkgreen"),
pT = c("T0" = "khaki","T1-T2" = "khaki", "T3-T4" ="brown2"),
pN = c("N0" = "cornflowerblue", "N1-N2" ="orange2"),
Grade = c("GX" = "grey","G1" = "coral", "G2" ="darkgreen", "G3" = "yellow3"),
ACT = c("TRUE" = "#C1211A", "FALSE" ="#008BCE"),
MSI = c("MSS" = "grey", "MSI-High" ="black"),
BRAF.V600E = c("WT" = "grey", "MUT" ="black"),
RAS = c("WT" = "grey", "MUT" ="black"),
ctDNA.MRD = c("POSITIVE" = "red3", "NEGATIVE" ="blue"),
ctDNA.Surveillance = c("POSITIVE" = "red3", "NEGATIVE" ="blue"),
ctDNA.anytime = c("POSITIVE" = "red3", "NEGATIVE" ="blue"),
DFS.Event = c("TRUE" = "red3", "FALSE" ="blue")
)
)
ht <- Heatmap(matrix(nrow = 0, ncol = length(circ_data$Stage)),show_row_names = FALSE,cluster_rows = F,cluster_columns = FALSE, top_annotation = ha)
pdf("heatmap.pdf",width = 7, height = 7)
draw(ht, annotation_legend_side = "bottom")
dev.off()#DFS by ctDNA at the MRD Window - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.MRD, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.MRD, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.MRD=NEGATIVE 328 29 NA NA NA
ctDNA.MRD=POSITIVE 62 36 11.3 8.67 NAevent_summary <- circ_data %>%
group_by(ctDNA.MRD) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.MRD, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA MRD window | All pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.MRD, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.MRD=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 110.000 25.000 0.903 0.019 0.859 0.935
ctDNA.MRD=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 12.0000 36.0000 0.3608 0.0669 0.2332 0.4898 circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.MRD, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.MRD, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.MRDPOSITIVE 2.2879 9.8539 0.2517 9.088 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.MRDPOSITIVE 9.854 0.1015 6.016 16.14
Concordance= 0.742 (se = 0.029 )
Likelihood ratio test= 74.36 on 1 df, p=<2e-16
Wald test = 82.6 on 1 df, p=<2e-16
Score (logrank) test = 124 on 1 df, p=<2e-16cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 9.85 (6.02-16.14); p = 0"#ctDNA sample positive in the MRD Window - 2-10 week intervals
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$MRD.Window == TRUE,]
circ_data <- circ_data[!is.na(circ_data$cfDNAconc), ]
filtered_data <- circ_data %>%
filter(ctDNA.MRD >= 2 & ctDNA.MRD <= 10) #intervals for 2-10 weeks
filtered_data$ctDNA_bucket <- cut(filtered_data$ctDNA.MRD,
breaks = c(2, 4, 6, 8, 10),
right = FALSE,
labels = c("2-4", "4-6", "6-8", "8-10"))
filtered_data <- filtered_data %>%
filter(!is.na(ctDNA_bucket))
rate_by_bucket <- filtered_data %>%
group_by(ctDNA_bucket) %>%
summarise(
n_total = n(), # Total number of patients in the bucket
n_positive = sum(biomarker_status == "POSITIVE"), # Number of positive cases
n_negative = sum(biomarker_status == "NEGATIVE"), # Number of negative cases
percentage_positive = mean(biomarker_status == "POSITIVE") * 100, # Positivity rate
percentage_negative = mean(biomarker_status == "NEGATIVE") * 100 # Negativity rate
)
overall_stats <- filtered_data %>%
summarise(
total_samples = n(),
total_positive = sum(biomarker_status == "POSITIVE"),
total_negative = sum(biomarker_status == "NEGATIVE"),
overall_percentage_positive = mean(biomarker_status == "POSITIVE") * 100
)
combined_results <- bind_rows(rate_by_bucket, overall_stats)
print(combined_results)
# Create the stacked bar plot for positivity and negativity rates by bucket
bar_midpoints <- barplot(
t(as.matrix(rate_by_bucket[, c("percentage_positive", "percentage_negative")])), # Transpose to get the correct format
names.arg = rate_by_bucket$ctDNA_bucket,
col = c("red", "blue"), # Colors: red for positive, blue for negative
main = '% ctDNA Positive and Negative Samples at the MRD Window',
xlab = 'Weeks from Surgery',
ylab = '% ctDNA Samples',
ylim = c(0, 100),
legend = c("% Positive", "% Negative"), # Adding a legend for clarification
args.legend = list(x = "topright")
)
par(new = TRUE)plot(bar_midpoints, rate_by_bucket$n_total, type = "b", col = "black", pch = 19, axes = FALSE, xlab = "", ylab = "", lwd = 2)
axis(side = 4) # Add the secondary y-axis on the rightmtext("Total Number of Samples", side = 4, line = 3) # Label for the secondary y-axis
text(bar_midpoints, rate_by_bucket$n_total + 3, labels = rate_by_bucket$n_total, col = "black", cex = 0.8)
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$MRD.Window == TRUE,]
circ_data <- circ_data[!is.na(circ_data$cfDNAconc), ]
filtered_data <- circ_data %>%
filter(ctDNA.MRD >= 0 & ctDNA.MRD <= 10) #intervals for 0-10 weeks
filtered_data$ctDNA_bucket <- cut(filtered_data$ctDNA.MRD,
breaks = c(0, 2, 4, 6, 8, 10),
right = FALSE,
labels = c("0-2","2-4", "4-6", "6-8", "8-10"))
filtered_data <- filtered_data %>%
filter(!is.na(ctDNA_bucket))
rate_by_bucket <- filtered_data %>%
group_by(ctDNA_bucket) %>%
summarise(
n_total = n(), # Total number of patients in the bucket
n_positive = sum(biomarker_status == "POSITIVE"), # Number of positive cases
n_negative = sum(biomarker_status == "NEGATIVE"), # Number of negative cases
percentage_positive = mean(biomarker_status == "POSITIVE") * 100, # Positivity rate
percentage_negative = mean(biomarker_status == "NEGATIVE") * 100 # Negativity rate
)
overall_stats <- filtered_data %>%
summarise(
total_samples = n(),
total_positive = sum(biomarker_status == "POSITIVE"),
total_negative = sum(biomarker_status == "NEGATIVE"),
overall_percentage_positive = mean(biomarker_status == "POSITIVE") * 100
)
combined_results <- bind_rows(rate_by_bucket, overall_stats)
print(combined_results)
# Create the stacked bar plot for positivity and negativity rates by bucket
bar_midpoints <- barplot(
t(as.matrix(rate_by_bucket[, c("percentage_positive", "percentage_negative")])), # Transpose to get the correct format
names.arg = rate_by_bucket$ctDNA_bucket,
col = c("red", "blue"), # Colors: red for positive, blue for negative
main = '% ctDNA Positive and Negative Samples at the MRD Window',
xlab = 'Weeks from Surgery',
ylab = '% ctDNA Samples',
ylim = c(0, 100),
legend = c("% Positive", "% Negative"), # Adding a legend for clarification
args.legend = list(x = "topright")
)
par(new = TRUE)plot(bar_midpoints, rate_by_bucket$n_total, type = "b", col = "black", pch = 19, axes = FALSE, xlab = "", ylab = "", lwd = 2)
axis(side = 4) # Add the secondary y-axis on the rightmtext("Total Number of Samples", side = 4, line = 3) # Label for the secondary y-axis
text(bar_midpoints, rate_by_bucket$n_total + 3, labels = rate_by_bucket$n_total, col = "black", cex = 0.8)
#Median number of timepoints in the MRD Window
rm(list=ls())
setwd("~/Downloads")
filtered_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
filtered_data <- filtered_data[filtered_data$MRD.Window==TRUE,]
filtered_data <- filtered_data[!is.na(filtered_data$cfDNAconc), ]
filtered_data <- filtered_data %>%
filter(ctDNA.MRD >= 2 & ctDNA.MRD <= 10) #intervals for 2-10 weeks
# Calculate the median number of ctDNA tests per patient
median_ctDNA_tests <- filtered_data %>%
group_by(pts_id) %>%
summarise(num_tests = n()) %>%
summarise(median_tests = median(num_tests))
ctDNA_stats <- filtered_data %>%
group_by(pts_id) %>%
tally() %>%
summarise(
median_tests = median(n),
min_tests = min(n),
max_tests = max(n)
)
print(median_ctDNA_tests)print(ctDNA_stats)#Multivariate cox regression for DFS at the MRD Window & Age threshold as 50 years - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"), labels = c("Negative", "Positive"))
circ_data$Gender <- factor(circ_data$Gender, levels = c("Female", "Male"))
circ_data$Age.Group2 <- factor(circ_data$Age.Group2, levels = c("1", "2"), labels = c("<50", "≥50"))
circ_data$PrimSite <- factor(circ_data$PrimSite, levels = c("Right-sided colon", "Left-sided colon"))
circ_data$pT <- factor(circ_data$pT, levels = c("T1-T2", "T3-T4"))
circ_data$pN <- factor(circ_data$pN, levels = c("N0", "N1-N2"))
circ_data$MSI <- factor(circ_data$MSI, levels = c("MSS", "MSI-High"))
circ_data$BRAF.V600E <- factor(circ_data$BRAF.V600E, levels = c("WT", "MUT"), labels = c("WT", "V600E"))
circ_data$RAS <- factor(circ_data$RAS, levels = c("WT", "MUT"), labels = c("WT", "Mut"))
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
cox_fit <- coxph(surv_object ~ ctDNA.MRD + Gender + Age.Group2 + PrimSite + pT + pN + MSI + BRAF.V600E + RAS, data=circ_data)
ggforest(cox_fit, data = circ_data, main = "Multivariate Regression Model for DFS", refLabel = "Reference Group")
test.ph <- cox.zph(cox_fit)
# Adjust p-values using False Discovery Rate (FDR) adjustment (Benjamini-Hochberg method)
p_values <- summary(cox_fit)$coefficients[, 5]
adjusted_p_values <- p.adjust(p_values, method = "fdr")
results <- data.frame(
Variable = rownames(summary(cox_fit)$coefficients),
Original_P_Value = p_values,
FDR_Adjusted_P_Value = adjusted_p_values
)
print(results)#Univariate cox regression for factors used in MVA - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$Gender <- factor(circ_data$Gender, levels = c("Female", "Male")) #univariate for gender
cox_fit <- coxph(surv_object ~ Gender, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ Gender, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
GenderMale 0.3259 1.3853 0.2550 1.278 0.201
exp(coef) exp(-coef) lower .95 upper .95
GenderMale 1.385 0.7219 0.8404 2.284
Concordance= 0.535 (se = 0.032 )
Likelihood ratio test= 1.67 on 1 df, p=0.2
Wald test = 1.63 on 1 df, p=0.2
Score (logrank) test = 1.65 on 1 df, p=0.2cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.39 (0.84-2.28); p = 0.201"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$Age.Group2 <- factor(circ_data$Age.Group2, levels = c("1", "2"), labels = c("<50", "≥50")) #univariate for Age
cox_fit <- coxph(surv_object ~ Age.Group2, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ Age.Group2, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
Age.Group2≥50 0.1934 1.2133 0.3102 0.623 0.533
exp(coef) exp(-coef) lower .95 upper .95
Age.Group2≥50 1.213 0.8242 0.6606 2.229
Concordance= 0.526 (se = 0.024 )
Likelihood ratio test= 0.4 on 1 df, p=0.5
Wald test = 0.39 on 1 df, p=0.5
Score (logrank) test = 0.39 on 1 df, p=0.5cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.21 (0.66-2.23); p = 0.533"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$PrimSite <- factor(circ_data$PrimSite, levels = c("Right-sided colon", "Left-sided colon")) #univariate for Tumor Location
cox_fit <- coxph(surv_object ~ PrimSite, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ PrimSite, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
PrimSiteLeft-sided colon 0.2250 1.2523 0.2482 0.907 0.365
exp(coef) exp(-coef) lower .95 upper .95
PrimSiteLeft-sided colon 1.252 0.7985 0.77 2.037
Concordance= 0.518 (se = 0.032 )
Likelihood ratio test= 0.82 on 1 df, p=0.4
Wald test = 0.82 on 1 df, p=0.4
Score (logrank) test = 0.83 on 1 df, p=0.4cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.25 (0.77-2.04); p = 0.365"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$pT <- factor(circ_data$pT, levels = c("T1-T2", "T3-T4")) #univariate for Overall T Stage
cox_fit <- coxph(surv_object ~ pT, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ pT, data = circ_data)
n= 386, number of events= 65
(4 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
pTT3-T4 1.0960 2.9923 0.5163 2.123 0.0338 *
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
pTT3-T4 2.992 0.3342 1.088 8.231
Concordance= 0.549 (se = 0.017 )
Likelihood ratio test= 6.26 on 1 df, p=0.01
Wald test = 4.51 on 1 df, p=0.03
Score (logrank) test = 4.98 on 1 df, p=0.03cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 2.99 (1.09-8.23); p = 0.034"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$pN <- factor(circ_data$pN, levels = c("N0", "N1-N2")) #univariate for Overall N Stage
cox_fit <- coxph(surv_object ~ pN, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ pN, data = circ_data)
n= 388, number of events= 65
(2 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
pNN1-N2 1.6966 5.4556 0.3592 4.723 2.32e-06 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
pNN1-N2 5.456 0.1833 2.698 11.03
Concordance= 0.667 (se = 0.023 )
Likelihood ratio test= 31.89 on 1 df, p=2e-08
Wald test = 22.31 on 1 df, p=2e-06
Score (logrank) test = 28.18 on 1 df, p=1e-07cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 5.46 (2.7-11.03); p = 0"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$MSI <- factor(circ_data$MSI, levels = c("MSS", "MSI-High")) #univariate for MSI
cox_fit <- coxph(surv_object ~ MSI, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ MSI, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
MSIMSI-High 0.2018 1.2236 0.3312 0.609 0.542
exp(coef) exp(-coef) lower .95 upper .95
MSIMSI-High 1.224 0.8173 0.6393 2.342
Concordance= 0.531 (se = 0.028 )
Likelihood ratio test= 0.35 on 1 df, p=0.6
Wald test = 0.37 on 1 df, p=0.5
Score (logrank) test = 0.37 on 1 df, p=0.5cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.22 (0.64-2.34); p = 0.542"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$BRAF.V600E <- factor(circ_data$BRAF.V600E, levels = c("WT", "MUT"), labels = c("WT", "V600E")) #univariate for BRAF
cox_fit <- coxph(surv_object ~ BRAF.V600E, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ BRAF.V600E, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
BRAF.V600EV600E 0.5464 1.7271 0.3199 1.708 0.0876 .
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
BRAF.V600EV600E 1.727 0.579 0.9226 3.233
Concordance= 0.552 (se = 0.028 )
Likelihood ratio test= 2.59 on 1 df, p=0.1
Wald test = 2.92 on 1 df, p=0.09
Score (logrank) test = 2.99 on 1 df, p=0.08cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.73 (0.92-3.23); p = 0.088"rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$RAS <- factor(circ_data$RAS, levels = c("WT", "MUT"), labels = c("WT", "Mut")) #univariate for RAS
cox_fit <- coxph(surv_object ~ RAS, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ RAS, data = circ_data)
n= 390, number of events= 65
coef exp(coef) se(coef) z Pr(>|z|)
RASMut 0.1691 1.1843 0.2507 0.675 0.5
exp(coef) exp(-coef) lower .95 upper .95
RASMut 1.184 0.8444 0.7246 1.936
Concordance= 0.511 (se = 0.031 )
Likelihood ratio test= 0.45 on 1 df, p=0.5
Wald test = 0.46 on 1 df, p=0.5
Score (logrank) test = 0.46 on 1 df, p=0.5cox_fit_summary <- summary(cox_fit)
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 1.18 (0.72-1.94); p = 0.5"#DFS by ctDNA at the MRD Window - Stage II
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "III")),]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.MRD, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.MRD, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.MRD=NEGATIVE 141 5 NA NA NA
ctDNA.MRD=POSITIVE 10 3 NA 11.1 NAevent_summary <- circ_data %>%
group_by(ctDNA.MRD) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.MRD, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA MRD window | Stage II", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.MRD, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.MRD=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 46.000 5.000 0.949 0.023 0.878 0.979
ctDNA.MRD=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 6.000 3.000 0.700 0.145 0.329 0.892 circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.MRD, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.MRD, data = circ_data)
n= 151, number of events= 8
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.MRDPOSITIVE 2.1194 8.3262 0.7317 2.897 0.00377 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.MRDPOSITIVE 8.326 0.1201 1.985 34.93
Concordance= 0.68 (se = 0.09 )
Likelihood ratio test= 6.3 on 1 df, p=0.01
Wald test = 8.39 on 1 df, p=0.004
Score (logrank) test = 12.02 on 1 df, p=5e-04cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 8.33 (1.98-34.93); p = 0.004"#DFS by ctDNA at the MRD Window - Stage III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "II")),]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.MRD, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.MRD, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.MRD=NEGATIVE 170 24 NA NA NA
ctDNA.MRD=POSITIVE 51 33 10 7.13 16.2event_summary <- circ_data %>%
group_by(ctDNA.MRD) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.MRD, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA MRD window | Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.MRD, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.MRD=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 57.0000 20.0000 0.8582 0.0303 0.7863 0.9073
ctDNA.MRD=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 6.0000 33.0000 0.2704 0.0714 0.1433 0.4146 circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.MRD, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.MRD, data = circ_data)
n= 221, number of events= 57
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.MRDPOSITIVE 2.0479 7.7515 0.2728 7.506 6.08e-14 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.MRDPOSITIVE 7.752 0.129 4.541 13.23
Concordance= 0.731 (se = 0.031 )
Likelihood ratio test= 53.81 on 1 df, p=2e-13
Wald test = 56.35 on 1 df, p=6e-14
Score (logrank) test = 77.35 on 1 df, p=<2e-16cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 7.75 (4.54-13.23); p = 0"#DFS by ctDNA at the MRD Window - Stage II & Risk Groups
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "III")),]
circ_data <- circ_data[circ_data$Risk.Group!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.Stage.II.Risk <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.Stage.II.Risk = case_when(
ctDNA.MRD == "NEGATIVE" & Risk.Group == "Low" ~ 1,
ctDNA.MRD == "POSITIVE" & Risk.Group == "Low" ~ 2,
ctDNA.MRD == "NEGATIVE" & Risk.Group == "High" ~ 3,
ctDNA.MRD == "POSITIVE" & Risk.Group == "High" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[!is.na(circ_data$ctDNA.Stage.II.Risk), ]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Stage.II.Risk, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Stage.II.Risk, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Stage.II.Risk=1 30 1 NA NA NA
ctDNA.Stage.II.Risk=2 2 0 NA NA NA
ctDNA.Stage.II.Risk=3 110 4 NA NA NA
ctDNA.Stage.II.Risk=4 8 3 NA 11.1 NAevent_summary <- circ_data %>%
group_by(ctDNA.Stage.II.Risk) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Stage.II.Risk, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA MRD & Stage II Risk Groups", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & Low Risk", "ctDNA(+) & Low Risk", "ctDNA(-) & High Risk", "ctDNA(+) & High Risk"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Stage.II.Risk, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Stage.II.Risk=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 11.0000 1.0000 0.9600 0.0392 0.7484 0.9943
ctDNA.Stage.II.Risk=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24 1 0 1 0 NA NA
ctDNA.Stage.II.Risk=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 35.0000 4.0000 0.9465 0.0278 0.8549 0.9809
ctDNA.Stage.II.Risk=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 5.000 3.000 0.625 0.171 0.229 0.861 circ_data$ctDNA.Stage.II.Risk <- factor(circ_data$ctDNA.Stage.II.Risk, levels=c("1","2","3","4"), labels = c("ctDNA(-) & Low Risk", "ctDNA(+) & Low Risk", "ctDNA(-) & High Risk", "ctDNA(+) & High Risk"))
cox_fit <- coxphf(surv_object ~ ctDNA.Stage.II.Risk, data=circ_data)
summary(cox_fit)coxphf(formula = surv_object ~ ctDNA.Stage.II.Risk, data = circ_data)
Model fitted by Penalized ML
Confidence intervals and p-values by Profile Likelihood
coef se(coef) exp(coef) lower 0.95 upper 0.95 Chisq p
ctDNA.Stage.II.RiskctDNA(+) & Low Risk 1.54746514 1.827452 4.6995424 0.03219659 88.15970 0.670679229 0.41281490
ctDNA.Stage.II.RiskctDNA(-) & High Risk -0.06180044 1.054866 0.9400705 0.17388788 9.38239 0.004245486 0.94804868
ctDNA.Stage.II.RiskctDNA(+) & High Risk 2.30998495 1.092244 10.0742730 1.65326183 104.09858 6.196660240 0.01279916
Likelihood ratio test=9.240477 on 3 df, p=0.02625872, n=150
Wald test = 9.971067 on 3 df, p = 0.01881368
Covariance-Matrix:
ctDNA.Stage.II.RiskctDNA(+) & Low Risk ctDNA.Stage.II.RiskctDNA(-) & High Risk ctDNA.Stage.II.RiskctDNA(+) & High Risk
ctDNA.Stage.II.RiskctDNA(+) & Low Risk 3.3395808 0.8333478 0.8321149
ctDNA.Stage.II.RiskctDNA(-) & High Risk 0.8333478 1.1127416 0.8335422
ctDNA.Stage.II.RiskctDNA(+) & High Risk 0.8321149 0.8335422 1.1929966#DFS by ctDNA at the MRD Window - Stage III & Risk Groups
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "II")),]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.Stage.III.Risk <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.Stage.III.Risk = case_when(
ctDNA.MRD == "NEGATIVE" & Risk.Group == "Low" ~ 1,
ctDNA.MRD == "POSITIVE" & Risk.Group == "Low" ~ 2,
ctDNA.MRD == "NEGATIVE" & Risk.Group == "High" ~ 3,
ctDNA.MRD == "POSITIVE" & Risk.Group == "High" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[!is.na(circ_data$ctDNA.Stage.III.Risk), ]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Stage.III.Risk, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Stage.III.Risk, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Stage.III.Risk=1 103 10 NA NA NA
ctDNA.Stage.III.Risk=2 13 9 12.9 6.14 NA
ctDNA.Stage.III.Risk=3 67 14 NA NA NA
ctDNA.Stage.III.Risk=4 38 24 9.0 6.01 NAevent_summary <- circ_data %>%
group_by(ctDNA.Stage.III.Risk) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Stage.III.Risk, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA MRD & Stage III Risk Groups", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & Low Risk", "ctDNA(+) & Low Risk", "ctDNA(-) & High Risk", "ctDNA(+) & High Risk"), legend.title="")
summary(KM_curve, times= c(18, 24))Call: survfit(formula = surv_object ~ ctDNA.Stage.III.Risk, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Stage.III.Risk=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18 57 8 0.909 0.0312 0.825 0.954
24 34 1 0.889 0.0367 0.791 0.942
ctDNA.Stage.III.Risk=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18.0000 2.0000 8.0000 0.3077 0.1417 0.0793 0.5780
ctDNA.Stage.III.Risk=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18 38 11 0.814 0.0512 0.688 0.893
24 23 0 0.814 0.0512 0.688 0.893
ctDNA.Stage.III.Risk=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18 8 24 0.304 0.0812 0.157 0.464
24 6 0 0.304 0.0812 0.157 0.464circ_data$ctDNA.Stage.III.Risk <- factor(circ_data$ctDNA.Stage.III.Risk, levels=c("1","2","3","4"), labels = c("ctDNA(-) & Low Risk", "ctDNA(+) & Low Risk", "ctDNA(-) & High Risk", "ctDNA(+) & High Risk"))
cox_fit <- coxph(surv_object ~ ctDNA.Stage.III.Risk, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Stage.III.Risk, data = circ_data)
n= 221, number of events= 57
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Stage.III.RiskctDNA(+) & Low Risk 2.4892 12.0518 0.4642 5.362 8.23e-08 ***
ctDNA.Stage.III.RiskctDNA(-) & High Risk 0.7524 2.1221 0.4141 1.817 0.0692 .
ctDNA.Stage.III.RiskctDNA(+) & High Risk 2.3908 10.9222 0.3791 6.307 2.85e-10 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Stage.III.RiskctDNA(+) & Low Risk 12.052 0.08297 4.8517 29.937
ctDNA.Stage.III.RiskctDNA(-) & High Risk 2.122 0.47123 0.9425 4.778
ctDNA.Stage.III.RiskctDNA(+) & High Risk 10.922 0.09156 5.1955 22.961
Concordance= 0.757 (se = 0.033 )
Likelihood ratio test= 57.24 on 3 df, p=2e-12
Wald test = 55.9 on 3 df, p=4e-12
Score (logrank) test = 79.25 on 3 df, p=<2e-16rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "II")),]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.Stage.III.Risk <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.Stage.III.Risk = case_when(
ctDNA.MRD == "NEGATIVE" & Risk.Group == "Low" ~ 1,
ctDNA.MRD == "POSITIVE" & Risk.Group == "Low" ~ 2,
ctDNA.MRD == "NEGATIVE" & Risk.Group == "High" ~ 3,
ctDNA.MRD == "POSITIVE" & Risk.Group == "High" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Stage.III.Risk, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Stage.III.Risk, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Stage.III.Risk=1 103 10 NA NA NA
ctDNA.Stage.III.Risk=2 13 9 12.9 6.14 NA
ctDNA.Stage.III.Risk=3 67 14 NA NA NA
ctDNA.Stage.III.Risk=4 38 24 9.0 6.01 NAsurv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Stage.III.Risk, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA MRD & Stage III Risk Groups", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & Low Risk", "ctDNA(+) & Low Risk", "ctDNA(-) & High Risk", "ctDNA(+) & High Risk"), legend.title="")
summary(KM_curve, times= c(18))Call: survfit(formula = surv_object ~ ctDNA.Stage.III.Risk, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Stage.III.Risk=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18.0000 57.0000 8.0000 0.9095 0.0312 0.8248 0.9543
ctDNA.Stage.III.Risk=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18.0000 2.0000 8.0000 0.3077 0.1417 0.0793 0.5780
ctDNA.Stage.III.Risk=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18.0000 38.0000 11.0000 0.8141 0.0512 0.6878 0.8932
ctDNA.Stage.III.Risk=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
18.0000 8.0000 24.0000 0.3038 0.0812 0.1574 0.4641 circ_data$ctDNA.Stage.III.Risk <- factor(circ_data$ctDNA.Stage.III.Risk, levels=c("2","4","1","3"))
cox_fit <- coxph(surv_object ~ ctDNA.Stage.III.Risk, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Stage.III.Risk, data = circ_data)
n= 221, number of events= 57
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Stage.III.Risk4 -0.09842 0.90627 0.39258 -0.251 0.802
ctDNA.Stage.III.Risk1 -2.48922 0.08297 0.46424 -5.362 8.23e-08 ***
ctDNA.Stage.III.Risk3 -1.73680 0.17608 0.43279 -4.013 5.99e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Stage.III.Risk4 0.90627 1.103 0.41985 1.9562
ctDNA.Stage.III.Risk1 0.08297 12.052 0.03340 0.2061
ctDNA.Stage.III.Risk3 0.17608 5.679 0.07539 0.4113
Concordance= 0.757 (se = 0.033 )
Likelihood ratio test= 57.24 on 3 df, p=2e-12
Wald test = 55.9 on 3 df, p=4e-12
Score (logrank) test = 79.25 on 3 df, p=<2e-16#OS by ctDNA at the MRD Window - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$OS.Event <- as.logical(circ_data$OS.Event)
circ_data$OS.months <- as.numeric(circ_data$OS.months)
circ_data$DFS.months=circ_data$OS.months-2.5
circ_data <- circ_data[circ_data$OS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
survfit(Surv(time = circ_data$OS.months, event = circ_data$OS.Event)~ctDNA.MRD, data = circ_data)Call: survfit(formula = Surv(time = circ_data$OS.months, event = circ_data$OS.Event) ~
ctDNA.MRD, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.MRD=NEGATIVE 334 8 NA NA NA
ctDNA.MRD=POSITIVE 71 7 40.4 40.4 NAevent_summary <- circ_data %>%
group_by(ctDNA.MRD) %>%
summarise(
Total = n(),
Events = sum(OS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$OS.months, event = circ_data$OS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.MRD, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="OS - ctDNA MRD window | All pts", ylab= "Overall Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.MRD, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.MRD=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
2.40e+01 1.52e+02 7.00e+00 9.75e-01 9.49e-03 9.48e-01 9.88e-01
ctDNA.MRD=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 30.0000 5.0000 0.9064 0.0408 0.7856 0.9607 circ_data$ctDNA.MRD <- factor(circ_data$ctDNA.MRD, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.MRD, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.MRD, data = circ_data)
n= 405, number of events= 15
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.MRDPOSITIVE 1.551 4.717 0.519 2.988 0.0028 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.MRDPOSITIVE 4.717 0.212 1.705 13.04
Concordance= 0.633 (se = 0.068 )
Likelihood ratio test= 7.93 on 1 df, p=0.005
Wald test = 8.93 on 1 df, p=0.003
Score (logrank) test = 10.85 on 1 df, p=0.001cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 4.72 (1.71-13.04); p = 0.003"#DFS by ctDNA at the Surveillance Window - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.Surveillance!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Surveillance, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Surveillance, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Surveillance=NEGATIVE 540 20 NA NA NA
ctDNA.Surveillance=POSITIVE 83 54 16.2 13.7 28.4event_summary <- circ_data %>%
group_by(ctDNA.Surveillance) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Surveillance, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA Surveillance window | All pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Surveillance, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Surveillance=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
2.40e+01 2.83e+02 1.60e+01 9.64e-01 9.06e-03 9.41e-01 9.78e-01
ctDNA.Surveillance=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 21.0000 44.0000 0.4087 0.0597 0.2916 0.5222 circ_data$ctDNA.Surveillance <- factor(circ_data$ctDNA.Surveillance, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.Surveillance, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Surveillance, data = circ_data)
n= 623, number of events= 74
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.SurveillancePOSITIVE 3.2926 26.9135 0.2645 12.45 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.SurveillancePOSITIVE 26.91 0.03716 16.03 45.19
Concordance= 0.817 (se = 0.026 )
Likelihood ratio test= 173.2 on 1 df, p=<2e-16
Wald test = 155 on 1 df, p=<2e-16
Score (logrank) test = 344 on 1 df, p=<2e-16cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 26.91 (16.03-45.19); p = 0"#DFS by ctDNA at the Surveillance Window - Stages II
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "III")),]
circ_data <- circ_data[circ_data$ctDNA.Surveillance!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Surveillance, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Surveillance, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Surveillance=NEGATIVE 198 6 NA NA NA
ctDNA.Surveillance=POSITIVE 16 9 21.2 11.1 NAevent_summary <- circ_data %>%
group_by(ctDNA.Surveillance) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Surveillance, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA Surveillance window | Stage II", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Surveillance, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Surveillance=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 99.000 5.000 0.967 0.015 0.920 0.986
ctDNA.Surveillance=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 3.000 7.000 0.480 0.149 0.187 0.725 circ_data$ctDNA.Surveillance <- factor(circ_data$ctDNA.Surveillance, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.Surveillance, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Surveillance, data = circ_data)
n= 214, number of events= 15
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.SurveillancePOSITIVE 3.4093 30.2449 0.5373 6.345 2.23e-10 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.SurveillancePOSITIVE 30.24 0.03306 10.55 86.7
Concordance= 0.782 (se = 0.063 )
Likelihood ratio test= 34.66 on 1 df, p=4e-09
Wald test = 40.26 on 1 df, p=2e-10
Score (logrank) test = 94.31 on 1 df, p=<2e-16cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 30.24 (10.55-86.7); p = 0"#DFS by ctDNA at the Surveillance Window - Stages III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[!(circ_data$Stage %in% c("I", "II")),]
circ_data <- circ_data[circ_data$ctDNA.Surveillance!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Surveillance, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Surveillance, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Surveillance=NEGATIVE 306 13 NA NA NA
ctDNA.Surveillance=POSITIVE 64 45 15.8 13.1 28.4event_summary <- circ_data %>%
group_by(ctDNA.Surveillance) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Surveillance, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA Surveillance window | Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Surveillance, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Surveillance=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 163.0000 10.0000 0.9610 0.0123 0.9281 0.9790
ctDNA.Surveillance=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 18.0000 37.0000 0.3800 0.0646 0.2553 0.5037 circ_data$ctDNA.Surveillance <- factor(circ_data$ctDNA.Surveillance, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.Surveillance, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Surveillance, data = circ_data)
n= 370, number of events= 58
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.SurveillancePOSITIVE 3.2208 25.0476 0.3173 10.15 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.SurveillancePOSITIVE 25.05 0.03992 13.45 46.65
Concordance= 0.827 (se = 0.027 )
Likelihood ratio test= 129.4 on 1 df, p=<2e-16
Wald test = 103 on 1 df, p=<2e-16
Score (logrank) test = 222.1 on 1 df, p=<2e-16cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 25.05 (13.45-46.65); p = 0"#OS by ctDNA at the Surveillance Window - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.Surveillance!="",]
circ_data$OS.Event <- as.logical(circ_data$OS.Event)
circ_data$OS.months <- as.numeric(circ_data$OS.months)
circ_data$DFS.months=circ_data$OS.months-2.5
circ_data <- circ_data[circ_data$OS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
survfit(Surv(time = circ_data$OS.months, event = circ_data$OS.Event)~ctDNA.Surveillance, data = circ_data)Call: survfit(formula = Surv(time = circ_data$OS.months, event = circ_data$OS.Event) ~
ctDNA.Surveillance, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Surveillance=NEGATIVE 540 4 NA NA NA
ctDNA.Surveillance=POSITIVE 83 4 NA NA NAevent_summary <- circ_data %>%
group_by(ctDNA.Surveillance) %>%
summarise(
Total = n(),
Events = sum(OS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$OS.months, event = circ_data$OS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Surveillance, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="OS - ctDNA Surveillance window | All pts", ylab= "Overall Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Surveillance, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Surveillance=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
2.40e+01 3.32e+02 3.00e+00 9.94e-01 3.65e-03 9.80e-01 9.98e-01
ctDNA.Surveillance=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 41.0000 3.0000 0.9594 0.0231 0.8785 0.9868 circ_data$ctDNA.Surveillance <- factor(circ_data$ctDNA.Surveillance, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.Surveillance, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Surveillance, data = circ_data)
n= 623, number of events= 8
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.SurveillancePOSITIVE 1.9570 7.0779 0.7088 2.761 0.00576 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.SurveillancePOSITIVE 7.078 0.1413 1.764 28.39
Concordance= 0.687 (se = 0.094 )
Likelihood ratio test= 6.66 on 1 df, p=0.01
Wald test = 7.62 on 1 df, p=0.006
Score (logrank) test = 10.36 on 1 df, p=0.001cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 7.08 (1.76-28.39); p = 0.006"#ctDNA sample positive in the Surveillance Window - 6mo intervals
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$Surveillance.Window == TRUE,]
circ_data <- circ_data[!is.na(circ_data$cfDNAconc), ]
circ_data$ctDNA_bucket <- cut(circ_data$ctDNA.Surv,
breaks = seq(0, max(circ_data$ctDNA.Surv, na.rm = TRUE), by = 6),
right = FALSE,
labels = paste0(seq(0, max(circ_data$ctDNA.Surv, na.rm = TRUE) - 6, by = 6),
"-",
seq(6, max(circ_data$ctDNA.Surv, na.rm = TRUE), by = 6)))
rate_by_bucket <- circ_data %>%
group_by(ctDNA_bucket) %>%
summarise(
n_total = n(), # Total number of patients in the bucket
n_positive = sum(biomarker_status == "POSITIVE"), # Number of positive cases
n_negative = sum(biomarker_status == "NEGATIVE"), # Number of negative cases
percentage_positive = mean(biomarker_status == "POSITIVE") * 100, # Positivity rate
percentage_negative = mean(biomarker_status == "NEGATIVE") * 100 # Negativity rate
)
overall_totals <- circ_data %>%
summarise(
ctDNA_bucket = "Total", # Label for the total row
n_total = n(), # Total number of patients across all buckets
n_positive = sum(biomarker_status == "POSITIVE"), # Total number of positive cases across all buckets
percentage_positive = mean(biomarker_status == "POSITIVE") * 100, # Overall positivity rate
n_negative = sum(biomarker_status == "NEGATIVE"), # Total number of negative cases across all buckets
percentage_negative = mean(biomarker_status == "NEGATIVE") * 100 # Overall negativity rate
)
positivity_rate_with_total <- bind_rows(rate_by_bucket, overall_totals)
print(positivity_rate_with_total)bar_midpoints <- barplot(
t(as.matrix(rate_by_bucket[, c("percentage_positive", "percentage_negative")])), # Transpose to get correct format
names.arg = rate_by_bucket$ctDNA_bucket,
col = c("red", "blue"), # Red for positivity and blue for negativity
main = '% ctDNA Positive and Negative Samples by 6-month Interval',
xlab = 'Time from Start of Surveillance (Months)',
ylab = '% ctDNA Samples',
ylim = c(0, 100),
legend = c("% Positive", "% Negative"), # Adding a legend for clarification
args.legend = list(x = "topright")
)
par(new = TRUE)plot(bar_midpoints, rate_by_bucket$n_total, type = "b", col = "black", pch = 19, axes = FALSE, xlab = "", ylab = "", lwd = 2)
axis(side = 4) # Add the secondary y-axis on the rightmtext("Total Number of Samples", side = 4, line = 3) # Label for the secondary y-axis
text(bar_midpoints, rate_by_bucket$n_total + 3, labels = rate_by_bucket$n_total, col = "black", cex = 0.8)
#Median number of timepoints in the Surveillance Window
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$Surveillance.Window == TRUE,]
circ_data <- circ_data[!is.na(circ_data$cfDNAconc), ]
# Calculate the median number of ctDNA tests per patient
ctDNA_stats <- circ_data %>%
group_by(pts_id) %>%
tally() %>%
summarise(
median_tests = median(n),
min_tests = min(n),
max_tests = max(n)
)
# Print the result
print(ctDNA_stats)#Time between imaging at the Surveillance window by ctDNA status
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC Imaging Frequency.csv")
tally(~ctDNA_Surveillance, data=circ_data, margins = TRUE)ctDNA_Surveillance
NEGATIVE POSITIVE Total
151 22 173 median_AverageDateDifference <- aggregate(AverageDateDifference ~ ctDNA_Surveillance, data = circ_data, FUN = median)
print(median_AverageDateDifference)circ_data$ctDNA_Surveillance <- factor(circ_data$ctDNA_Surveillance, levels = c("NEGATIVE", "POSITIVE"), labels = c("Negative (n=151)","Positive (n=22)"))
boxplot(AverageDateDifference~ctDNA_Surveillance, data=circ_data, main="Average Time between Imaging at Surveillance Window", xlab="ctDNA at the Surveillance Window", ylab="Average Time Between Imaging (Days)", col="white",border="black")
m1<-wilcox.test(AverageDateDifference ~ ctDNA_Surveillance, data=circ_data, na.rm=TRUE, exact=FALSE, conf.int=TRUE)
print(m1)
Wilcoxon rank sum test with continuity correction
data: AverageDateDifference by ctDNA_Surveillance
W = 2327, p-value = 6.322e-06
alternative hypothesis: true location shift is not equal to 0
95 percent confidence interval:
106.25 269.00
sample estimates:
difference in location
181 rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC Imaging Frequency.csv")
tally(~ctDNA_Surveillance, data=circ_data, margins = TRUE)ctDNA_Surveillance
NEGATIVE POSITIVE Total
151 22 173 median_MedianDateDifference <- aggregate(MedianDateDifference ~ ctDNA_Surveillance, data = circ_data, FUN = median)
print(median_MedianDateDifference)circ_data$ctDNA_Surveillance <- factor(circ_data$ctDNA_Surveillance, levels = c("NEGATIVE", "POSITIVE"), labels = c("Negative (n=151)","Positive (n=22)"))
boxplot(MedianDateDifference~ctDNA_Surveillance, data=circ_data, main="Median Time between Imaging at Surveillance Window", xlab="ctDNA at the Surveillance Window", ylab="Median Time Between Imaging (Days)", col="white",border="black")
m2<-wilcox.test(MedianDateDifference ~ ctDNA_Surveillance, data=circ_data, na.rm=TRUE, exact=FALSE, conf.int=TRUE)
print(m2)
Wilcoxon rank sum test with continuity correction
data: MedianDateDifference by ctDNA_Surveillance
W = 2336.5, p-value = 5.008e-06
alternative hypothesis: true location shift is not equal to 0
95 percent confidence interval:
90.00009 269.00006
sample estimates:
difference in location
158.7506 #Time-dependent analysis in surveillance window - all stages
rm(list=ls())
setwd("~/Downloads")
dt_final <- read.csv("CLIA CRC Time_Dependent_analysis.csv")
dt_final <- dt_final[dt_final$CRC.Cohort=="TRUE",]
dt_final <- dt_final[dt_final$tstart!="",]
datatable(dt_final, filter = "top")
# Syntax if there is not time-dependent covariate
# fit <- coxph(Surv(dfs_time, dfs_event) ~ biomarker_status,
# data = dt_final)
# summary(fit)
fit <- coxph(Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)Warning: Stop time must be > start time, NA createdsummary(fit)Call:
coxph(formula = Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)
n= 3456, number of events= 95
(1211 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
biomarker_statusPOSITIVE 3.6369 37.9742 0.2235 16.27 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
biomarker_statusPOSITIVE 37.97 0.02633 24.51 58.84
Concordance= 0.816 (se = 0.024 )
Likelihood ratio test= 256.4 on 1 df, p=<2e-16
Wald test = 264.9 on 1 df, p=<2e-16
Score (logrank) test = 679.7 on 1 df, p=<2e-16#Time-dependent analysis in surveillance window - all stages ACT-treated
rm(list=ls())
setwd("~/Downloads")
dt_final <- read.csv("CLIA CRC Time_Dependent_analysis.csv")
dt_final <- dt_final[dt_final$CRC.Cohort=="TRUE",]
dt_final <- dt_final[dt_final$ACT=="TRUE",]
dt_final <- dt_final[dt_final$tstart!="",]
datatable(dt_final, filter = "top")
# Syntax if there is not time-dependent covariate
# fit <- coxph(Surv(dfs_time, dfs_event) ~ biomarker_status,
# data = dt_final)
# summary(fit)
fit <- coxph(Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)Warning: Stop time must be > start time, NA createdsummary(fit)Call:
coxph(formula = Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)
n= 2178, number of events= 69
(1014 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
biomarker_statusPOSITIVE 3.5174 33.6971 0.2601 13.52 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
biomarker_statusPOSITIVE 33.7 0.02968 20.24 56.11
Concordance= 0.807 (se = 0.029 )
Likelihood ratio test= 175.9 on 1 df, p=<2e-16
Wald test = 182.8 on 1 df, p=<2e-16
Score (logrank) test = 443.9 on 1 df, p=<2e-16#Time-dependent analysis in surveillance window - all stages Non-ACT-treated
rm(list=ls())
setwd("~/Downloads")
dt_final <- read.csv("CLIA CRC Time_Dependent_analysis.csv")
dt_final <- dt_final[dt_final$CRC.Cohort=="TRUE",]
dt_final <- dt_final[dt_final$ACT=="FALSE",]
dt_final <- dt_final[dt_final$tstart!="",]
datatable(dt_final, filter = "top")
# Syntax if there is not time-dependent covariate
# fit <- coxph(Surv(dfs_time, dfs_event) ~ biomarker_status,
# data = dt_final)
# summary(fit)
fit <- coxph(Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)Warning: Stop time must be > start time, NA createdsummary(fit)Call:
coxph(formula = Surv(tstart, tstop, dfs_event) ~ biomarker_status,
data = dt_final)
n= 1278, number of events= 26
(197 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
biomarker_statusPOSITIVE 3.8078 45.0513 0.4362 8.73 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
biomarker_statusPOSITIVE 45.05 0.0222 19.16 105.9
Concordance= 0.835 (se = 0.043 )
Likelihood ratio test= 76.31 on 1 df, p=<2e-16
Wald test = 76.21 on 1 df, p=<2e-16
Score (logrank) test = 214.3 on 1 df, p=<2e-16#DFS by ctDNA subgroups at the Surveillance window
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Surveillance.Groups!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~Surveillance.Groups, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
Surveillance.Groups, data = circ_data)
n events median 0.95LCL 0.95UCL
Surveillance.Groups=All Negative 540 20 NA NA NA
Surveillance.Groups=All Positive 35 30 8.05 6.9 13.6
Surveillance.Groups=Negative-Positive 48 24 31.04 21.3 NAevent_summary <- circ_data %>%
group_by(Surveillance.Groups) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$Surveillance.Groups <- factor(circ_data$Surveillance.Groups, levels=c("All Negative","Negative-Positive","All Positive"))
KM_curve <- survfit(surv_object ~ Surveillance.Groups, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","purple", "red"), title="DFS - ctDNA Subgroups Surveillance Window | All Stages", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("All Negative","Negative-Positive","All Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ Surveillance.Groups, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
Surveillance.Groups=All Negative
time n.risk n.event survival std.err lower 95% CI upper 95% CI
2.40e+01 2.83e+02 1.60e+01 9.64e-01 9.06e-03 9.41e-01 9.78e-01
Surveillance.Groups=Negative-Positive
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 17.0000 17.0000 0.5803 0.0801 0.4086 0.7182
Surveillance.Groups=All Positive
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 4.0000 27.0000 0.1657 0.0702 0.0576 0.3222 cox_fit <- coxph(surv_object ~ Surveillance.Groups, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ Surveillance.Groups, data = circ_data)
n= 623, number of events= 74
coef exp(coef) se(coef) z Pr(>|z|)
Surveillance.GroupsNegative-Positive 2.7806 16.1288 0.3041 9.144 <2e-16 ***
Surveillance.GroupsAll Positive 4.2247 68.3543 0.3033 13.929 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
Surveillance.GroupsNegative-Positive 16.13 0.06200 8.887 29.27
Surveillance.GroupsAll Positive 68.35 0.01463 37.721 123.86
Concordance= 0.833 (se = 0.027 )
Likelihood ratio test= 198.8 on 2 df, p=<2e-16
Wald test = 196.3 on 2 df, p=<2e-16
Score (logrank) test = 536.6 on 2 df, p=<2e-16rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Surveillance.Groups!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
circ_data$Surveillance.Groups <- factor(circ_data$Surveillance.Groups, levels=c("Negative-Positive","All Positive", "All Negative"))
cox_fit <- coxph(surv_object ~ Surveillance.Groups, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ Surveillance.Groups, data = circ_data)
n= 623, number of events= 74
coef exp(coef) se(coef) z Pr(>|z|)
Surveillance.GroupsAll Positive 1.4441 4.2380 0.2826 5.111 3.2e-07 ***
Surveillance.GroupsAll Negative -2.7806 0.0620 0.3041 -9.144 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
Surveillance.GroupsAll Positive 4.238 0.236 2.43591 7.3734
Surveillance.GroupsAll Negative 0.062 16.129 0.03416 0.1125
Concordance= 0.833 (se = 0.027 )
Likelihood ratio test= 198.8 on 2 df, p=<2e-16
Wald test = 196.3 on 2 df, p=<2e-16
Score (logrank) test = 536.6 on 2 df, p=<2e-16#Demographics table for ctDNA positive at the Surveillance window with no radiological recurrence
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.Surveillance=="POSITIVE",]
circ_data <- circ_data[circ_data$DFS.Event==FALSE,]
circ_data_subset <- circ_data %>%
select(
Age,
Gender,
PrimSite,
pT,
pN,
Stage,
Grade,
NAC,
ACT,
MSI,
BRAF.V600E,
RAS,
OS.months) %>%
mutate(
Age = as.numeric(Age),
Gender = factor(Gender, levels = c("Male", "Female")),
PrimSite = factor(PrimSite, levels = c("Right-sided colon", "Left-sided colon")),
pT = factor(pT, levels = c("T0","T1-T2", "T3-T4")),
pN = factor(pN, levels = c("N0", "N1-N2")),
Stage = factor(Stage, levels = c("I","II", "III")),
Grade = factor(Grade, levels = c("G1", "G2", "G3","GX")),
NAC = factor(NAC, levels = c("TRUE", "FALSE"), labels = c("Neoadjuvant Chemotherapy", "Upfront Surgery")),
ACT = factor(ACT, levels = c("TRUE", "FALSE"), labels = c("Adjuvant Chemotherapy", "Observation")),
MSI = factor(MSI, levels = c("MSS", "MSI-High")),
BRAF.V600E = factor(BRAF.V600E, levels = c("WT", "MUT"), labels = c("BRAF WT", "BRAF V600E")),
RAS = factor(RAS, levels = c("WT", "MUT"), labels = c("RAS WT", "RAS Mut")),
OS.months = as.numeric(OS.months))
table1 <- circ_data_subset %>%
tbl_summary(
statistic = list(
all_continuous() ~ "{median} ({min} - {max})",
all_categorical() ~ "{n} ({p}%)")) %>%
bold_labels()
table1| Characteristic | N = 291 |
|---|---|
| Age | 57 (28 - 86) |
| Gender | |
| Male | 12 (41%) |
| Female | 17 (59%) |
| PrimSite | |
| Right-sided colon | 12 (41%) |
| Left-sided colon | 17 (59%) |
| pT | |
| T0 | 0 (0%) |
| T1-T2 | 4 (14%) |
| T3-T4 | 25 (86%) |
| pN | |
| N0 | 9 (31%) |
| N1-N2 | 20 (69%) |
| Stage | |
| I | 3 (10%) |
| II | 7 (24%) |
| III | 19 (66%) |
| Grade | |
| G1 | 4 (14%) |
| G2 | 19 (66%) |
| G3 | 6 (21%) |
| GX | 0 (0%) |
| NAC | |
| Neoadjuvant Chemotherapy | 0 (0%) |
| Upfront Surgery | 29 (100%) |
| ACT | |
| Adjuvant Chemotherapy | 19 (66%) |
| Observation | 10 (34%) |
| MSI | |
| MSS | 26 (90%) |
| MSI-High | 3 (10%) |
| BRAF.V600E | |
| BRAF WT | 26 (90%) |
| BRAF V600E | 3 (10%) |
| RAS | |
| RAS WT | 17 (59%) |
| RAS Mut | 12 (41%) |
| OS.months | 22 (7 - 62) |
| 1 Median (Range); n (%) | |
fit1 <- as_flex_table(
table1,
include = everything(),
return_calls = FALSE,
strip_md_bold = TRUE)
fit1Characteristic | N = 291 |
|---|---|
Age | 57 (28 - 86) |
Gender | |
Male | 12 (41%) |
Female | 17 (59%) |
PrimSite | |
Right-sided colon | 12 (41%) |
Left-sided colon | 17 (59%) |
pT | |
T0 | 0 (0%) |
T1-T2 | 4 (14%) |
T3-T4 | 25 (86%) |
pN | |
N0 | 9 (31%) |
N1-N2 | 20 (69%) |
Stage | |
I | 3 (10%) |
II | 7 (24%) |
III | 19 (66%) |
Grade | |
G1 | 4 (14%) |
G2 | 19 (66%) |
G3 | 6 (21%) |
GX | 0 (0%) |
NAC | |
Neoadjuvant Chemotherapy | 0 (0%) |
Upfront Surgery | 29 (100%) |
ACT | |
Adjuvant Chemotherapy | 19 (66%) |
Observation | 10 (34%) |
MSI | |
MSS | 26 (90%) |
MSI-High | 3 (10%) |
BRAF.V600E | |
BRAF WT | 26 (90%) |
BRAF V600E | 3 (10%) |
RAS | |
RAS WT | 17 (59%) |
RAS Mut | 12 (41%) |
OS.months | 22 (7 - 62) |
1Median (Range); n (%) | |
save_as_docx(fit1, path= "~/Downloads/table1.docx")#DFS by ctDNA at 12 months post-surgery
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.12months!="",]
circ_data$DFS.months=circ_data$DFS.months-12
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.12months, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.12months, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.12months=NEGATIVE 136 6 NA NA NA
ctDNA.12months=POSITIVE 14 7 6.69 3.74 NAevent_summary <- circ_data %>%
group_by(ctDNA.12months) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.12months, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA at 12 months | All pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.12months, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.12months=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 31.0000 6.0000 0.9365 0.0263 0.8593 0.9720
ctDNA.12months=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 1.000 7.000 0.373 0.156 0.105 0.650 circ_data$ctDNA.12months <- factor(circ_data$ctDNA.12months, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.12months, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.12months, data = circ_data)
n= 150, number of events= 13
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.12monthsPOSITIVE 2.8501 17.2889 0.5708 4.993 5.94e-07 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.12monthsPOSITIVE 17.29 0.05784 5.648 52.92
Concordance= 0.748 (se = 0.068 )
Likelihood ratio test= 20.99 on 1 df, p=5e-06
Wald test = 24.93 on 1 df, p=6e-07
Score (logrank) test = 45.91 on 1 df, p=1e-11cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 17.29 (5.65-52.92); p = 0"#DFS by ctDNA 4-weeks post-adjuvant chemotherapy
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.postACT!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.postACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.postACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.postACT=NEGATIVE 112 17 NA NA NA
ctDNA.postACT=POSITIVE 6 4 7.54 3.45 NAevent_summary <- circ_data %>%
group_by(ctDNA.postACT) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.postACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","red"), title="DFS - ctDNA 4 weeks post-ACT | All pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA Negative", "ctDNA Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.postACT, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.postACT=NEGATIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 42.000 14.000 0.845 0.039 0.749 0.906
ctDNA.postACT=POSITIVE
time n.risk n.event survival std.err lower 95% CI upper 95% CIcirc_data$ctDNA.postACT <- factor(circ_data$ctDNA.postACT, levels=c("NEGATIVE","POSITIVE"))
cox_fit <- coxph(surv_object ~ ctDNA.postACT, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.postACT, data = circ_data)
n= 118, number of events= 21
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.postACTPOSITIVE 2.4960 12.1334 0.5794 4.308 1.65e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.postACTPOSITIVE 12.13 0.08242 3.897 37.77
Concordance= 0.612 (se = 0.051 )
Likelihood ratio test= 11.57 on 1 df, p=7e-04
Wald test = 18.56 on 1 df, p=2e-05
Score (logrank) test = 30.09 on 1 df, p=4e-08cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 12.13 (3.9-37.77); p = 0"#DFS by ACT treatment in MRD negative - High Risk Stage II or Stage III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$ctDNA.MRD=="NEGATIVE",]
circ_data$DFS.months=circ_data$DFS.months-2
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ACT=FALSE 93 5 NA NA NA
ACT=TRUE 187 23 NA NA NAevent_summary <- circ_data %>%
group_by(ACT) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("red","blue"), title="DFS - ctDNA MRD Negative ACT vs Observation | High Risk Stage II or Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Observation", "ACT"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ACT, data = circ_data, conf.int = 0.95,
conf.type = "log-log")
ACT=FALSE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 30.0000 5.0000 0.9186 0.0372 0.8054 0.9673
ACT=TRUE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 68.000 19.000 0.878 0.027 0.813 0.921 circ_data$ACT <- factor(circ_data$ACT, levels=c("TRUE","FALSE"))
cox_fit <- coxph(surv_object ~ ACT, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ACT, data = circ_data)
n= 280, number of events= 28
coef exp(coef) se(coef) z Pr(>|z|)
ACTFALSE -0.7813 0.4578 0.4935 -1.583 0.113
exp(coef) exp(-coef) lower .95 upper .95
ACTFALSE 0.4578 2.184 0.174 1.204
Concordance= 0.571 (se = 0.04 )
Likelihood ratio test= 2.93 on 1 df, p=0.09
Wald test = 2.51 on 1 df, p=0.1
Score (logrank) test = 2.64 on 1 df, p=0.1cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 0.46 (0.17-1.2); p = 0.113"#Adjusted HR "ACT vs no ACT" - age, gender, MSI and pathological stage
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$ctDNA.MRD=="NEGATIVE",]
circ_data$DFS.months=circ_data$DFS.months-2
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data$ACT <- factor(circ_data$ACT, levels=c("TRUE","FALSE"))
circ_data$Age.Group <- factor(circ_data$Age.Group, levels = c("1", "2"), labels = c("<70", "≥70"))
circ_data$Gender <- factor(circ_data$Gender, levels = c("Female", "Male"))
circ_data$MSI <- factor(circ_data$MSI, levels = c("MSS", "MSI-High"))
circ_data$Stage <- factor(circ_data$Stage, levels = c("II", "III"))
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
cox_fit <- coxph(surv_object ~ ACT + Age.Group + Gender + MSI + Stage, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ACT + Age.Group + Gender + MSI +
Stage, data = circ_data)
n= 280, number of events= 28
coef exp(coef) se(coef) z Pr(>|z|)
ACTFALSE -0.1808 0.8346 0.6123 -0.295 0.7678
Age.Group≥70 0.3185 1.3750 0.4494 0.709 0.4785
GenderMale 0.5723 1.7724 0.4009 1.428 0.1534
MSIMSI-High 0.2271 1.2550 0.5780 0.393 0.6944
StageIII 1.3172 3.7330 0.6395 2.060 0.0394 *
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ACTFALSE 0.8346 1.1982 0.2513 2.772
Age.Group≥70 1.3750 0.7273 0.5699 3.318
GenderMale 1.7724 0.5642 0.8078 3.889
MSIMSI-High 1.2550 0.7968 0.4042 3.896
StageIII 3.7330 0.2679 1.0658 13.075
Concordance= 0.668 (se = 0.055 )
Likelihood ratio test= 10.58 on 5 df, p=0.06
Wald test = 8.62 on 5 df, p=0.1
Score (logrank) test = 9.59 on 5 df, p=0.09# Adjust p-values using False Discovery Rate (FDR) adjustment (Benjamini-Hochberg method)
p_values <- summary(cox_fit)$coefficients[, 5]
adjusted_p_values <- p.adjust(p_values, method = "fdr")
results <- data.frame(
Variable = rownames(summary(cox_fit)$coefficients),
Original_P_Value = p_values,
FDR_Adjusted_P_Value = adjusted_p_values
)
print(results)#DFS by ACT treatment in MRD positive - High Risk Stage II or Stage III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$ctDNA.MRD=="POSITIVE",]
circ_data$DFS.months=circ_data$DFS.months-2
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ACT=FALSE 10 9 2.41 1.22 NA
ACT=TRUE 51 29 13.38 10.19 NAevent_summary <- circ_data %>%
group_by(ACT) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("red","blue"), title="DFS - ctDNA MRD Positive ACT vs Observation | High Risk Stage II or Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Observation", "ACT"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ACT, data = circ_data, conf.int = 0.95,
conf.type = "log-log")
ACT=FALSE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.00000 1.00000 9.00000 0.10000 0.09487 0.00572 0.35813
ACT=TRUE
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 11.0000 29.0000 0.3632 0.0744 0.2218 0.5060 circ_data$ACT <- factor(circ_data$ACT, levels=c("TRUE","FALSE"))
cox_fit <- coxph(surv_object ~ ACT, data=circ_data)
ggforest(cox_fit,data = circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ACT, data = circ_data)
n= 61, number of events= 38
coef exp(coef) se(coef) z Pr(>|z|)
ACTFALSE 1.3968 4.0422 0.3909 3.573 0.000353 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ACTFALSE 4.042 0.2474 1.879 8.697
Concordance= 0.613 (se = 0.036 )
Likelihood ratio test= 9.94 on 1 df, p=0.002
Wald test = 12.77 on 1 df, p=4e-04
Score (logrank) test = 14.86 on 1 df, p=1e-04cox_fit_summary <- summary(cox_fit)
# Extract values for HR, 95% CI, and p-value
HR <- cox_fit_summary$coefficients[2]
lower_CI <- cox_fit_summary$conf.int[3]
upper_CI <- cox_fit_summary$conf.int[4]
p_value <- cox_fit_summary$coefficients[5]
label_text <- paste0("HR = ", round(HR, 2), " (", round(lower_CI, 2), "-", round(upper_CI, 2), "); p = ", round(p_value, 3))
print(label_text)[1] "HR = 4.04 (1.88-8.7); p = 0"#Adjusted HR "ACT vs no ACT" - age, gender, MSI and pathological stage
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$ctDNA.MRD=="POSITIVE",]
circ_data$DFS.months=circ_data$DFS.months-2
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_datadf <- as.data.frame(circ_data)
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data$ACT <- factor(circ_data$ACT, levels=c("TRUE","FALSE"))
circ_data$Age.Group <- factor(circ_data$Age.Group, levels = c("1", "2"), labels = c("<70", "≥70"))
circ_data$Gender <- factor(circ_data$Gender, levels = c("Female", "Male"))
circ_data$MSI <- factor(circ_data$MSI, levels = c("MSS", "MSI-High"))
circ_data$Stage <- factor(circ_data$Stage, levels = c("II", "III"))
surv_object <- Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
cox_fit <- coxph(surv_object ~ ACT + Age.Group + Gender + Stage, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ACT + Age.Group + Gender + Stage,
data = circ_data)
n= 61, number of events= 38
coef exp(coef) se(coef) z Pr(>|z|)
ACTFALSE 1.9310 6.8967 0.4580 4.216 2.48e-05 ***
Age.Group≥70 0.5599 1.7505 0.3603 1.554 0.1202
GenderMale 0.2086 1.2320 0.3471 0.601 0.5478
StageIII 1.3374 3.8093 0.5942 2.251 0.0244 *
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ACTFALSE 6.897 0.1450 2.8106 16.923
Age.Group≥70 1.750 0.5713 0.8640 3.547
GenderMale 1.232 0.8117 0.6239 2.432
StageIII 3.809 0.2625 1.1887 12.208
Concordance= 0.669 (se = 0.044 )
Likelihood ratio test= 18.74 on 4 df, p=9e-04
Wald test = 20.61 on 4 df, p=4e-04
Score (logrank) test = 22.94 on 4 df, p=1e-04# Adjust p-values using False Discovery Rate (FDR) adjustment (Benjamini-Hochberg method)
p_values <- summary(cox_fit)$coefficients[, 5]
adjusted_p_values <- p.adjust(p_values, method = "fdr")
results <- data.frame(
Variable = rownames(summary(cox_fit)$coefficients),
Original_P_Value = p_values,
FDR_Adjusted_P_Value = adjusted_p_values
)
print(results)#DFS by ctDNA at the MRD Window & ACT - MSS Stable all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$MSI=="MSS",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.ACT <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.ACT = case_when(
ctDNA.MRD == "NEGATIVE" & ACT == "TRUE" ~ 1,
ctDNA.MRD == "POSITIVE" & ACT == "TRUE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ACT == "FALSE" ~ 3,
ctDNA.MRD == "POSITIVE" & ACT == "FALSE" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.ACT=1 169 19 NA NA NA
ctDNA.ACT=2 48 27 13.07 9.69 NA
ctDNA.ACT=3 106 6 NA NA NA
ctDNA.ACT=4 3 2 7.42 5.29 NAevent_summary <- circ_data %>%
group_by(ctDNA.ACT) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple","red"), title="DFS - ctDNA MRD & ACT | MSS pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.ACT, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.ACT=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 61.000 15.000 0.892 0.027 0.825 0.934
ctDNA.ACT=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 9.0000 27.0000 0.3640 0.0773 0.2174 0.5120
ctDNA.ACT=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 37.000 6.000 0.921 0.032 0.829 0.965
ctDNA.ACT=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.00000 1.00000 2.00000 0.33333 0.27217 0.00896 0.77415 circ_data$ctDNA.ACT <- factor(circ_data$ctDNA.ACT, levels=c("1","2","3","4"), labels = c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"))
cox_fit <- coxph(surv_object ~ ctDNA.ACT, data=circ_data) #modify maxexit to reveal NA values in cox_fit
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.ACT, data = circ_data)
n= 326, number of events= 54
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.ACTctDNA(+) & ACT 2.0237 7.5661 0.3021 6.700 2.09e-11 ***
ctDNA.ACTctDNA(-) & No ACT -0.6820 0.5056 0.4683 -1.456 0.14536
ctDNA.ACTctDNA(+) & No ACT 2.1276 8.3950 0.7463 2.851 0.00436 **
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.ACTctDNA(+) & ACT 7.5661 0.1322 4.1856 13.677
ctDNA.ACTctDNA(-) & No ACT 0.5056 1.9778 0.2019 1.266
ctDNA.ACTctDNA(+) & No ACT 8.3950 0.1191 1.9443 36.247
Concordance= 0.759 (se = 0.034 )
Likelihood ratio test= 61 on 3 df, p=4e-13
Wald test = 65.55 on 3 df, p=4e-14
Score (logrank) test = 98.87 on 3 df, p=<2e-16rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$MSI=="MSS",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.ACT <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.ACT = case_when(
ctDNA.MRD == "NEGATIVE" & ACT == "TRUE" ~ 1,
ctDNA.MRD == "POSITIVE" & ACT == "TRUE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ACT == "FALSE" ~ 3,
ctDNA.MRD == "POSITIVE" & ACT == "FALSE" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.ACT=1 169 19 NA NA NA
ctDNA.ACT=2 48 27 13.07 9.69 NA
ctDNA.ACT=3 106 6 NA NA NA
ctDNA.ACT=4 3 2 7.42 5.29 NAsurv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple","red"), title="DFS - ctDNA MRD & ACT | MSS pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.ACT, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.ACT=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 61.000 15.000 0.892 0.027 0.825 0.934
ctDNA.ACT=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 9.0000 27.0000 0.3640 0.0773 0.2174 0.5120
ctDNA.ACT=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 37.000 6.000 0.921 0.032 0.829 0.965
ctDNA.ACT=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.00000 1.00000 2.00000 0.33333 0.27217 0.00896 0.77415 circ_data$ctDNA.ACT <- factor(circ_data$ctDNA.ACT, levels=c("3","1","2","4"), labels = c("ctDNA(-) & No ACT","ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(+) & No ACT"))
cox_fit <- coxph(surv_object ~ ctDNA.ACT, data=circ_data)
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.ACT, data = circ_data)
n= 326, number of events= 54
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.ACTctDNA(-) & ACT 0.6820 1.9778 0.4683 1.456 0.145361
ctDNA.ACTctDNA(+) & ACT 2.7056 14.9638 0.4533 5.969 2.39e-09 ***
ctDNA.ACTctDNA(+) & No ACT 2.8096 16.6032 0.8192 3.430 0.000605 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.ACTctDNA(-) & ACT 1.978 0.50562 0.7898 4.953
ctDNA.ACTctDNA(+) & ACT 14.964 0.06683 6.1544 36.383
ctDNA.ACTctDNA(+) & No ACT 16.603 0.06023 3.3331 82.704
Concordance= 0.759 (se = 0.034 )
Likelihood ratio test= 61 on 3 df, p=4e-13
Wald test = 65.55 on 3 df, p=4e-14
Score (logrank) test = 98.87 on 3 df, p=<2e-16#DFS by ctDNA at the MRD Window & ACT - MSI High all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$MSI=="MSI-High",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.ACT <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.ACT = case_when(
ctDNA.MRD == "NEGATIVE" & ACT == "TRUE" ~ 1,
ctDNA.MRD == "POSITIVE" & ACT == "TRUE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ACT == "FALSE" ~ 3,
ctDNA.MRD == "POSITIVE" & ACT == "FALSE" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.ACT=1 19 4 NA NA NA
ctDNA.ACT=2 5 2 NA 3.09 NA
ctDNA.ACT=3 34 0 NA NA NA
ctDNA.ACT=4 6 5 1.91 1.41 NAevent_summary <- circ_data %>%
group_by(ctDNA.ACT) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple","red"), title="DFS - ctDNA MRD & ACT | MSI-High pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"), legend.title="")
summary(KM_curve, times= c(6, 24))Call: survfit(formula = surv_object ~ ctDNA.ACT, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.ACT=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6 15 3 0.839 0.0854 0.579 0.945
24 3 1 0.774 0.1003 0.502 0.910
ctDNA.ACT=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6 3 2 0.6 0.219 0.126 0.882
24 2 0 0.6 0.219 0.126 0.882
ctDNA.ACT=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6 31 0 1 0 NA NA
24 9 0 1 0 NA NA
ctDNA.ACT=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6.00000 1.00000 5.00000 0.16667 0.15215 0.00772 0.51680 circ_data$ctDNA.ACT <- factor(circ_data$ctDNA.ACT, levels=c("3","1","2","4"), labels = c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"))
cox_fit <- coxphf(surv_object ~ ctDNA.ACT, data=circ_data, maxit = 500, maxstep = 1) #modify maxexit to reveal NA values in cox_fit
summary(cox_fit)coxphf(formula = surv_object ~ ctDNA.ACT, data = circ_data, maxit = 500,
maxstep = 1)
Model fitted by Penalized ML
Confidence intervals and p-values by Profile Likelihood
coef se(coef) exp(coef) lower 0.95 upper 0.95 Chisq p
ctDNA.ACTctDNA(+) & ACT 2.813027 1.622557 16.66027 1.778749 2207.805 6.609234 1.014513e-02
ctDNA.ACTctDNA(-) & No ACT 3.712830 1.686348 40.96957 3.329543 5652.756 8.497597 3.556157e-03
ctDNA.ACTctDNA(+) & No ACT 4.942760 1.627985 140.15657 14.930527 18714.894 24.151928 8.902699e-07
Likelihood ratio test=26.64366 on 3 df, p=6.992091e-06, n=64
Wald test = 14.30441 on 3 df, p = 0.002518759
Covariance-Matrix:
ctDNA.ACTctDNA(+) & ACT ctDNA.ACTctDNA(-) & No ACT ctDNA.ACTctDNA(+) & No ACT
ctDNA.ACTctDNA(+) & ACT 2.632691 2.370004 2.371709
ctDNA.ACTctDNA(-) & No ACT 2.370004 2.843769 2.373722
ctDNA.ACTctDNA(+) & No ACT 2.371709 2.373722 2.650336rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data <- circ_data[circ_data$MSI=="MSI-High",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data$ctDNA.ACT <- NA #first we create the variable for the ctDNA & NAC combination, and we assign values
circ_data <- circ_data %>%
mutate(ctDNA.ACT = case_when(
ctDNA.MRD == "NEGATIVE" & ACT == "TRUE" ~ 1,
ctDNA.MRD == "POSITIVE" & ACT == "TRUE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ACT == "FALSE" ~ 3,
ctDNA.MRD == "POSITIVE" & ACT == "FALSE" ~ 4
))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.ACT, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.ACT, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.ACT=1 19 4 NA NA NA
ctDNA.ACT=2 5 2 NA 3.09 NA
ctDNA.ACT=3 34 0 NA NA NA
ctDNA.ACT=4 6 5 1.91 1.41 NAsurv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.ACT, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple","red"), title="DFS - ctDNA MRD & ACT | MSI-High pts", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("ctDNA(-) & ACT", "ctDNA(+) & ACT", "ctDNA(-) & No ACT", "ctDNA(+) & No ACT"), legend.title="")
summary(KM_curve, times= c(6))Call: survfit(formula = surv_object ~ ctDNA.ACT, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.ACT=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6.0000 15.0000 3.0000 0.8388 0.0854 0.5788 0.9451
ctDNA.ACT=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6.000 3.000 2.000 0.600 0.219 0.126 0.882
ctDNA.ACT=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6 31 0 1 0 NA NA
ctDNA.ACT=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
6.00000 1.00000 5.00000 0.16667 0.15215 0.00772 0.51680 circ_data$ctDNA.ACT <- factor(circ_data$ctDNA.ACT, levels=c("2","4","1","3"))
cox_fit <- coxphf(surv_object ~ ctDNA.ACT, data=circ_data, maxit = 500) #modify maxexit to reveal NA values in cox_fit
summary(cox_fit)coxphf(formula = surv_object ~ ctDNA.ACT, data = circ_data, maxit = 500)
Model fitted by Penalized ML
Confidence intervals and p-values by Profile Likelihood
coef se(coef) exp(coef) lower 0.95 upper 0.95 Chisq p
ctDNA.ACT4 1.2299304 0.8640952 3.42099155 0.771944824 20.3070261 2.600353 0.106839899
ctDNA.ACT1 -0.8998028 0.8581680 0.40664983 0.090118999 2.3336235 1.163885 0.280661276
ctDNA.ACT3 -3.7128298 1.6863479 0.02440835 0.000176905 0.3003415 8.497597 0.003556157
Likelihood ratio test=26.64366 on 3 df, p=6.992091e-06, n=64
Wald test = 14.30441 on 3 df, p = 0.002518759
Covariance-Matrix:
ctDNA.ACT4 ctDNA.ACT1 ctDNA.ACT3
ctDNA.ACT4 0.7466605 0.4717519 0.4700468
ctDNA.ACT1 0.4717519 0.7364524 0.4737650
ctDNA.ACT3 0.4700468 0.4737650 2.8437691#DFS by ctDNA Dynamics from MRD to Surveillance Window - all stages
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data <- circ_data %>%
mutate(ctDNA.Dynamics = case_when(
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 1,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "POSITIVE" ~ 3,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "POSITIVE" ~ 4
)) %>%
filter(!is.na(ctDNA.Dynamics))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[circ_data$ctDNA.Dynamics!="",]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Dynamics, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Dynamics, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Dynamics=1 263 10 NA NA NA
ctDNA.Dynamics=2 16 1 NA NA NA
ctDNA.Dynamics=3 25 11 28.4 15.77 NA
ctDNA.Dynamics=4 20 15 10.6 8.05 NAevent_summary <- circ_data %>%
group_by(ctDNA.Dynamics) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Dynamics, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA Dynamics from MRD to Surveillance Window | All Stages", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Persistently Negative", "Converted Negative","Converted Positive", "Persistently Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Dynamics, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Dynamics=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 93.000 8.000 0.961 0.014 0.922 0.981
ctDNA.Dynamics=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 8.0000 1.0000 0.9000 0.0949 0.4730 0.9853
ctDNA.Dynamics=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 6.000 9.000 0.519 0.125 0.260 0.727
ctDNA.Dynamics=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 2.0000 15.0000 0.2400 0.0980 0.0821 0.4428 circ_data$ctDNA.Dynamics <- factor(circ_data$ctDNA.Dynamics, levels=c("1","2","3","4"))
cox_fit <- coxph(surv_object ~ ctDNA.Dynamics, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Dynamics, data = circ_data)
n= 324, number of events= 37
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Dynamics2 0.5229 1.6869 1.0493 0.498 0.618
ctDNA.Dynamics3 2.6136 13.6480 0.4383 5.963 2.48e-09 ***
ctDNA.Dynamics4 3.5381 34.4018 0.4141 8.545 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Dynamics2 1.687 0.59281 0.2157 13.19
ctDNA.Dynamics3 13.648 0.07327 5.7804 32.22
ctDNA.Dynamics4 34.402 0.02907 15.2803 77.45
Concordance= 0.822 (se = 0.04 )
Likelihood ratio test= 79.02 on 3 df, p=<2e-16
Wald test = 77.71 on 3 df, p=<2e-16
Score (logrank) test = 168.8 on 3 df, p=<2e-16rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data <- circ_data %>%
mutate(ctDNA.Dynamics = case_when(
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 1,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "POSITIVE" ~ 3,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "POSITIVE" ~ 4
)) %>%
filter(!is.na(ctDNA.Dynamics))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[circ_data$ctDNA.Dynamics!="",]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Dynamics, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Dynamics, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Dynamics=1 263 10 NA NA NA
ctDNA.Dynamics=2 16 1 NA NA NA
ctDNA.Dynamics=3 25 11 28.4 15.77 NA
ctDNA.Dynamics=4 20 15 10.6 8.05 NAsurv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Dynamics, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA Dynamics from MRD to Surveillance Window | All Stages", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Persistently Negative", "Converted Negative","Converted Positive", "Persistently Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Dynamics, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Dynamics=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 93.000 8.000 0.961 0.014 0.922 0.981
ctDNA.Dynamics=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 8.0000 1.0000 0.9000 0.0949 0.4730 0.9853
ctDNA.Dynamics=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 6.000 9.000 0.519 0.125 0.260 0.727
ctDNA.Dynamics=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 2.0000 15.0000 0.2400 0.0980 0.0821 0.4428 circ_data$ctDNA.Dynamics <- factor(circ_data$ctDNA.Dynamics, levels=c("3","4","1","2"))
cox_fit <- coxph(surv_object ~ ctDNA.Dynamics, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Dynamics, data = circ_data)
n= 324, number of events= 37
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Dynamics4 0.92452 2.52065 0.39952 2.314 0.0207 *
ctDNA.Dynamics1 -2.61359 0.07327 0.43834 -5.963 2.48e-09 ***
ctDNA.Dynamics2 -2.09071 0.12360 1.04642 -1.998 0.0457 *
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Dynamics4 2.52065 0.3967 1.15197 5.515
ctDNA.Dynamics1 0.07327 13.6480 0.03103 0.173
ctDNA.Dynamics2 0.12360 8.0907 0.01590 0.961
Concordance= 0.822 (se = 0.04 )
Likelihood ratio test= 79.02 on 3 df, p=<2e-16
Wald test = 77.71 on 3 df, p=<2e-16
Score (logrank) test = 168.8 on 3 df, p=<2e-16#DFS by ctDNA Dynamics from MRD to Surveillance Window - High Risk Stage II or Stage III
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data <- circ_data %>%
mutate(ctDNA.Dynamics = case_when(
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 1,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "POSITIVE" ~ 3,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "POSITIVE" ~ 4
)) %>%
filter(!is.na(ctDNA.Dynamics))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[circ_data$ctDNA.Dynamics!="",]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Dynamics, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Dynamics, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Dynamics=1 224 10 NA NA NA
ctDNA.Dynamics=2 14 1 NA NA NA
ctDNA.Dynamics=3 22 10 28.4 16.03 NA
ctDNA.Dynamics=4 19 15 10.1 8.05 NAevent_summary <- circ_data %>%
group_by(ctDNA.Dynamics) %>%
summarise(
Total = n(),
Events = sum(DFS.Event),
Fraction = Events / n(),
Percentage = (Events / n()) * 100
)
print(event_summary)surv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Dynamics, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA Dynamics from MRD to Surveillance Window | High Risk Stage II or Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Persistently Negative", "Converted Negative","Converted Positive", "Persistently Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Dynamics, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Dynamics=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 77.0000 8.0000 0.9539 0.0166 0.9073 0.9774
ctDNA.Dynamics=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 7.000 1.000 0.889 0.105 0.433 0.984
ctDNA.Dynamics=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 6.000 8.000 0.542 0.128 0.272 0.750
ctDNA.Dynamics=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 2.0000 15.0000 0.1974 0.0948 0.0551 0.4032 circ_data$ctDNA.Dynamics <- factor(circ_data$ctDNA.Dynamics, levels=c("1","2","3","4"))
cox_fit <- coxph(surv_object ~ ctDNA.Dynamics, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Dynamics, data = circ_data)
n= 279, number of events= 36
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Dynamics2 0.4503 1.5688 1.0494 0.429 0.668
ctDNA.Dynamics3 2.4212 11.2599 0.4483 5.401 6.64e-08 ***
ctDNA.Dynamics4 3.4366 31.0803 0.4147 8.286 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Dynamics2 1.569 0.63743 0.2006 12.27
ctDNA.Dynamics3 11.260 0.08881 4.6765 27.11
ctDNA.Dynamics4 31.080 0.03217 13.7872 70.06
Concordance= 0.816 (se = 0.04 )
Likelihood ratio test= 72.14 on 3 df, p=1e-15
Wald test = 72.35 on 3 df, p=1e-15
Score (logrank) test = 150.9 on 3 df, p=<2e-16rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC_Clinical Data 122023.csv")
circ_data <- circ_data[circ_data$CLIA.CRC==TRUE,]
circ_data <- circ_data[circ_data$Risk.Stage==TRUE,]
circ_data <- circ_data[circ_data$ctDNA.MRD!="",]
circ_data$DFS.months=circ_data$DFS.months-2.5
circ_data <- circ_data[circ_data$DFS.months>=0,]
circ_data <- circ_data %>%
mutate(ctDNA.Dynamics = case_when(
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 1,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "NEGATIVE" ~ 2,
ctDNA.MRD == "NEGATIVE" & ctDNA.Surveillance == "POSITIVE" ~ 3,
ctDNA.MRD == "POSITIVE" & ctDNA.Surveillance == "POSITIVE" ~ 4
)) %>%
filter(!is.na(ctDNA.Dynamics))
circ_data$DFS.Event <- as.logical(circ_data$DFS.Event)
circ_data$DFS.months <- as.numeric(circ_data$DFS.months)
circ_data <- circ_data[circ_data$ctDNA.Dynamics!="",]
survfit(Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)~ctDNA.Dynamics, data = circ_data)Call: survfit(formula = Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event) ~
ctDNA.Dynamics, data = circ_data)
n events median 0.95LCL 0.95UCL
ctDNA.Dynamics=1 224 10 NA NA NA
ctDNA.Dynamics=2 14 1 NA NA NA
ctDNA.Dynamics=3 22 10 28.4 16.03 NA
ctDNA.Dynamics=4 19 15 10.1 8.05 NAsurv_object <-Surv(time = circ_data$DFS.months, event = circ_data$DFS.Event)
KM_curve <- survfit(surv_object ~ ctDNA.Dynamics, data = circ_data,conf.int=0.95,conf.type="log-log")
ggsurvplot(KM_curve, data = circ_data, pval = FALSE, conf.int = FALSE, risk.table = TRUE, break.time.by=6, palette=c("blue","green","purple", "red"), title="DFS - ctDNA Dynamics from MRD to Surveillance Window | High Risk Stage II or Stage III", ylab= "Disease-Free Survival", xlab="Time from Landmark Time point (Months)", legend.labs=c("Persistently Negative", "Converted Negative","Converted Positive", "Persistently Positive"), legend.title="")
summary(KM_curve, times= c(24))Call: survfit(formula = surv_object ~ ctDNA.Dynamics, data = circ_data,
conf.int = 0.95, conf.type = "log-log")
ctDNA.Dynamics=1
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 77.0000 8.0000 0.9539 0.0166 0.9073 0.9774
ctDNA.Dynamics=2
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 7.000 1.000 0.889 0.105 0.433 0.984
ctDNA.Dynamics=3
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.000 6.000 8.000 0.542 0.128 0.272 0.750
ctDNA.Dynamics=4
time n.risk n.event survival std.err lower 95% CI upper 95% CI
24.0000 2.0000 15.0000 0.1974 0.0948 0.0551 0.4032 circ_data$ctDNA.Dynamics <- factor(circ_data$ctDNA.Dynamics, levels=c("3","4","1","2"))
cox_fit <- coxph(surv_object ~ ctDNA.Dynamics, data=circ_data)
ggforest(cox_fit,data = circ_data) 
summary(cox_fit)Call:
coxph(formula = surv_object ~ ctDNA.Dynamics, data = circ_data)
n= 279, number of events= 36
coef exp(coef) se(coef) z Pr(>|z|)
ctDNA.Dynamics4 1.01533 2.76027 0.41237 2.462 0.0138 *
ctDNA.Dynamics1 -2.42124 0.08881 0.44832 -5.401 6.64e-08 ***
ctDNA.Dynamics2 -1.97094 0.13933 1.05059 -1.876 0.0607 .
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
ctDNA.Dynamics4 2.76027 0.3623 1.23012 6.1938
ctDNA.Dynamics1 0.08881 11.2599 0.03689 0.2138
ctDNA.Dynamics2 0.13933 7.1774 0.01777 1.0922
Concordance= 0.816 (se = 0.04 )
Likelihood ratio test= 72.14 on 3 df, p=1e-15
Wald test = 72.35 on 3 df, p=1e-15
Score (logrank) test = 150.9 on 3 df, p=<2e-16#Median cfDNA concentration across different windows post-surgery
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$cfDNA.Include == TRUE,]
circ_data$Timepoint <- factor(circ_data$Timepoint, levels = c("0-2 weeks", "2-4 weeks", "4-6 weeks", "6-8 weeks", "8-10 weeks","On-treatment", "Surveillance"))
circ_data$cfDNAconc <- as.numeric(circ_data$cfDNAconc)
median_cfDNAconc <- circ_data %>%
group_by(Timepoint) %>%
dplyr::summarize(
median_cfDNAconc = median(cfDNAconc, na.rm = TRUE),
n_observations = n(),
.groups = "drop"
)
print(median_cfDNAconc)boxplot(cfDNAconc~Timepoint, data=circ_data, main="median cfDNA Concentration", xlab="Intervals from Surgery", ylab="cfDNA Concentration", col="white",border="black", ylim = c(0, 40))
#Pairwise Wilcoxon-test p values
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$cfDNA.Include == TRUE,]
circ_data$Timepoint <- factor(circ_data$Timepoint, levels = c("0-2 weeks", "2-4 weeks", "4-6 weeks", "6-8 weeks", "8-10 weeks","On-treatment", "Surveillance"))
circ_data$cfDNAconc <- as.numeric(circ_data$cfDNAconc)
# Initialize a matrix to store p-values for manual Wilcoxon tests
timepoints <- levels(circ_data$Timepoint)
p_value_matrix <- matrix(NA, nrow = length(timepoints), ncol = length(timepoints))
rownames(p_value_matrix) <- timepoints
colnames(p_value_matrix) <- timepoints
# Perform pairwise Wilcoxon tests manually and store p-values
for (i in 1:length(timepoints)) {
for (j in i:length(timepoints)) {
if (i != j) {
# Subset data for the two Timepoints
data1 <- circ_data %>% filter(Timepoint == timepoints[i]) %>% pull(cfDNAconc)
data2 <- circ_data %>% filter(Timepoint == timepoints[j]) %>% pull(cfDNAconc)
# Perform Wilcoxon test and store the p-value
test_result <- wilcox.test(data1, data2, exact = FALSE)
p_value_matrix[i, j] <- test_result$p.value
p_value_matrix[j, i] <- test_result$p.value # Symmetric matrix
} else {
p_value_matrix[i, j] <- 1 # Set diagonal to 1 for clarity
}
}
}
# Replace NA values with 1 in the p-value matrix for completeness
p_value_matrix[is.na(p_value_matrix)] <- 1.00
# Melt the p-value matrix for plotting
p_value_data <- melt(p_value_matrix)
colnames(p_value_data) <- c("Timepoint1", "Timepoint2", "p_value")
# Add asterisks based on p-value thresholds
p_value_data <- p_value_data %>%
mutate(
asterisk = case_when(
p_value < 0.001 ~ "***",
p_value < 0.01 ~ "**",
p_value < 0.05 ~ "*",
TRUE ~ ""
)
)
# Create the heatmap with asterisks for significance levels
ggplot(p_value_data, aes(x = Timepoint1, y = Timepoint2, fill = p_value)) +
geom_tile(color = "white") + # Background tiles for grid
geom_text(aes(label = asterisk), color = "black", vjust = -0.5) + # Add asterisks
scale_fill_gradient(low = "lightgreen", high = "red") + # Gradient for p-values
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
labs(title = "Pairwise Wilcoxon-Test P-Values",
fill = "P-Value")#ctDNA positivity across different windows post-surgery
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$cfDNA.Include == TRUE,]
circ_data$Timepoint <- factor(circ_data$Timepoint, levels = c("0-2 weeks", "2-4 weeks", "4-6 weeks", "6-8 weeks", "8-10 weeks","On-treatment", "Surveillance"))
circ_data$cfDNAconc <- as.numeric(circ_data$cfDNAconc)
rate_by_timepoint <- circ_data %>%
group_by(Timepoint) %>%
summarise(
n_total = n(), # Total number of patients in each Timepoint
n_positive = sum(biomarker_status == "POSITIVE"), # Number of positive cases
n_negative = sum(biomarker_status == "NEGATIVE"), # Number of negative cases
percentage_positive = mean(biomarker_status == "POSITIVE") * 100, # Positivity rate
percentage_negative = mean(biomarker_status == "NEGATIVE") * 100 # Negativity rate
)
print(rate_by_timepoint)
# Create the stacked bar plot for positivity and negativity rates by Timepoint
bar_midpoints <- barplot(
t(as.matrix(rate_by_timepoint[, c("percentage_positive", "percentage_negative")])), # Transpose to get the correct format
names.arg = rate_by_timepoint$Timepoint,
col = c("red", "blue"), # Colors: red for positive, blue for negative
main = '% ctDNA Positive and Negative Samples by Timepoint',
xlab = 'Timepoint',
ylab = '% ctDNA Samples',
ylim = c(0, 100),
legend = c("% Positive", "% Negative"), # Adding a legend for clarification
args.legend = list(x = "topright")
)
par(new = TRUE)plot(bar_midpoints, rate_by_timepoint$n_total, type = "b", col = "black", pch = 19, axes = FALSE, xlab = "", ylab = "", lwd = 2)
axis(side = 4) # Add the secondary y-axis on the rightmtext("Total Number of Samples", side = 4, line = 3) # Label for the secondary y-axis
text(bar_midpoints, rate_by_timepoint$n_total + 3, labels = rate_by_timepoint$n_total, col = "black", cex = 0.8)
#Perform fisher's exact test matrix and heatmap
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$cfDNA.Include == TRUE,]
circ_data$Timepoint <- factor(circ_data$Timepoint, levels = c("0-2 weeks", "2-4 weeks", "4-6 weeks", "6-8 weeks", "8-10 weeks","On-treatment", "Surveillance"))
circ_data$cfDNAconc <- as.numeric(circ_data$cfDNAconc)
timepoints <- levels(circ_data$Timepoint) # Use ordered levels
p_value_matrix <- matrix(NA, nrow = length(timepoints), ncol = length(timepoints))
rownames(p_value_matrix) <- timepoints
colnames(p_value_matrix) <- timepoints
for (i in 1:length(timepoints)) {
for (j in i:length(timepoints)) {
if (i != j) {
# Subset data for the two Timepoints
subset1 <- circ_data %>% filter(Timepoint == timepoints[i])
subset2 <- circ_data %>% filter(Timepoint == timepoints[j])
# Create contingency table
contingency_table <- matrix(c(
sum(subset1$biomarker_status == "POSITIVE"), sum(subset1$biomarker_status == "NEGATIVE"),
sum(subset2$biomarker_status == "POSITIVE"), sum(subset2$biomarker_status == "NEGATIVE")
), nrow = 2, byrow = TRUE)
# Perform Fisher's exact test
test_result <- fisher.test(contingency_table)
# Store p-value in the matrix
p_value_matrix[i, j] <- test_result$p.value
p_value_matrix[j, i] <- test_result$p.value # Symmetric matrix
} else {
p_value_matrix[i, j] <- 1 # Set diagonal to 1 for clarity
}
}
}
print(p_value_matrix) 0-2 weeks 2-4 weeks 4-6 weeks 6-8 weeks 8-10 weeks On-treatment Surveillance
0-2 weeks 1.0000000 1.000000000 6.919353e-01 2.857000e-01 0.6738454083 1.000000e+00 4.643687e-01
2-4 weeks 1.0000000 1.000000000 1.588616e-01 1.035825e-02 0.2594894746 1.000000e+00 2.706764e-03
4-6 weeks 0.6919353 0.158861559 1.000000e+00 1.913544e-01 1.0000000000 5.680924e-02 8.035880e-09
6-8 weeks 0.2857000 0.010358245 1.913544e-01 1.000000e+00 0.4193440867 2.126873e-03 1.724838e-09
8-10 weeks 0.6738454 0.259489475 1.000000e+00 4.193441e-01 1.0000000000 1.498403e-01 3.034774e-04
On-treatment 1.0000000 1.000000000 5.680924e-02 2.126873e-03 0.1498402610 1.000000e+00 6.108761e-08
Surveillance 0.4643687 0.002706764 8.035880e-09 1.724838e-09 0.0003034774 6.108761e-08 1.000000e+00p_value_data <- melt(p_value_matrix)
colnames(p_value_data) <- c("Timepoint1", "Timepoint2", "p_value")
p_value_data <- p_value_data %>%
mutate(
asterisk = case_when(
p_value < 0.001 ~ "***",
p_value < 0.01 ~ "**",
p_value < 0.05 ~ "*",
TRUE ~ ""
)
)
# Create the heatmap with asterisks for significance levels
ggplot(p_value_data, aes(x = Timepoint1, y = Timepoint2, fill = p_value)) +
geom_tile(color = "white") + # Background tiles for grid
geom_text(aes(label = asterisk), color = "black", vjust = -0.5) + # Add asterisks
scale_fill_gradient(low = "lightgreen", high = "red") + # Gradient for p-values
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
labs(title = "Pairwise Fisher's Exact Test P-Values",
x = "Timepoint 1",
y = "Timepoint 2",
fill = "P-Value")
#Logistic regression for ctDNA positivity
rm(list=ls())
setwd("~/Downloads")
circ_data <- read.csv("CLIA CRC cfDNA data_ctDNA timepoints.csv")
circ_data <- circ_data[circ_data$cfDNA.Include == TRUE,]
# Calculate the 20th, 40th, 60th, 80th, and 100th percentiles for cfDNAconc
thresholds <- quantile(circ_data$cfDNAconc, probs = seq(0, 1, 0.2), na.rm = TRUE)
print(thresholds) 0% 20% 40% 60% 80% 100%
1.125000 3.566842 4.764706 6.244382 9.357955 265.185000 circ_data <- circ_data %>%
mutate(cfDNAlevels = cut(cfDNAconc,
breaks = thresholds,
include.lowest = TRUE,
labels = c("Q1", "Q2", "Q3", "Q4", "Q5")))
circ_data$cfDNAconc <- as.numeric(circ_data$cfDNAconc)
circ_data$biomarker_status <- factor(circ_data$biomarker_status, levels = c("NEGATIVE", "POSITIVE"))
circ_data$cfDNAlevels <- factor(circ_data$cfDNAlevels, levels = c("Q3", "Q1", "Q2", "Q4", "Q5"), labels = c("4.76-6.24 ng/mL (q3)", "<3.57 ng/mL (q1)", "3.57-4.76 ng/mL (q2)", "6.24-9.36 ng/mL (q4)", ">9.36 ng/mL (q5)"))
circ_data$Timepoint <- factor(circ_data$Timepoint, levels = c("4-6 weeks", "0-2 weeks", "2-4 weeks", "6-8 weeks", "8-10 weeks","On-treatment", "Surveillance"))
circ_data$Stage <- factor(circ_data$Stage, levels = c("I", "II", "III"))
circ_data$MSI <- factor(circ_data$MSI, levels = c("MSS", "MSI-High"))
circ_data$BRAF.V600E <- factor(circ_data$BRAF.V600E, levels = c("WT", "MUT"), labels = c("WT", "V600E"))
circ_data$RAS <- factor(circ_data$RAS, levels = c("WT", "MUT"), labels = c("WT", "Mut"))
logistic_model <- glm(biomarker_status ~ cfDNAlevels + Timepoint + Stage + MSI + BRAF.V600E + RAS,
data = circ_data,
family = binomial)
# Extract odds ratios and confidence intervals
results <- tidy(logistic_model, conf.int = TRUE, exponentiate = TRUE) %>%
mutate(term = factor(term, levels = term)) # Preserve term ordering
ggplot(results, aes(x = term, y = estimate, ymin = conf.low, ymax = conf.high)) +
geom_pointrange() +
geom_hline(yintercept = 1, linetype = "dashed", color = "gray") + # Reference line at OR = 1
coord_flip() +
labs(title = "Odds Ratios for Biomarker Status (Forest Plot)",
x = "Predictor",
y = "Odds Ratio (95% CI)") +
theme_minimal()
cox_model <- coxph(Surv(rep(1, nrow(circ_data)), biomarker_status == "POSITIVE") ~ cfDNAlevels + Timepoint + Stage + MSI + BRAF.V600E + RAS,
data = circ_data)
ggforest(cox_model, data = circ_data, main = "Odds Ratios for Biomarker Status", cpositions = c(0.02, 0.22, 0.4))
# Adjust p-values using False Discovery Rate (FDR) adjustment (Benjamini-Hochberg method)
p_values <- summary(cox_model)$coefficients[, 5]
adjusted_p_values <- p.adjust(p_values, method = "fdr")
results <- data.frame(
Variable = rownames(summary(cox_model)$coefficients),
Original_P_Value = p_values,
FDR_Adjusted_P_Value = adjusted_p_values
)
print(results)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